Mit
Mit
Mit
INSnnE
OF TECHNOLOGY
at the
Massachusetts Institute of Technology
LIBARIEJUN
2 8 2005
LIBRARIES
June 2005
02005 Massachusetts Institute of Technology, All rights reserved.
Certified by..
--
------------ t-----------------------------------------------------------------
/ John E. Fernandez
Associate Professor of Architecture
Thesis Supervisor
1
Certified by
mJ-Cnr
Jerome J. Connor
Professor of Civil and Environmental Engineering
Thesis Supervisor
Accepted by
Adele Naudd Santos
Chair, Committee on Graduate Students
Acting Head, Department of Architecture
Dean, School of Architecture and Planning
ARCHIVES
DYNAMIC INTERRELATIONSHIP
BETWEEN TECHNOLOGY AND ARCHITECTURE IN TALL BUILDINGS
by
Kyoung-Sun Moon
Thesis Committee:
John E. Fernandez
Associate Professor of Architecture
Department of Architecture
Massachusetts Institute of Technology
Jerome J. Connor
Professor of Civil and Environmental Engineering
Department of Civil & Environmental Engineering
Massachusetts Institute of Technology
Michael K. Kim
Professor of Architecture
Department of Architecture
University of Illinois at Urbana-Champaign
Abstract
The interrelationship between the technology and architecture of tall buildings is
investigated from the emergence of tall buildings in the late 1 9 th century to the present.
Through the historical research, a filtering concept is developed - original technology
and remedial technology - through which one can clearly understand the interrelationship
between the technological evolution and architectural esthetic and further stylistic
transition of tall buildings. More desirable visions for the future can be constructed
based on this concept.
Contemporary design practice of tall buildings is reviewed, and design guidelines are
provided for new design trends. Investigated in depth are the behavioral characteristics
and design methodology for diagrid structures, which emerge as a new direction in the
design of tall buildings with their powerful structural rationale and symbolic architectural
expression. Moreover, new technologies for tall building structures and facades are
developed for performance enhancement through design integration, and their
architectural potentials are explored. Special emphasis is placed on the research on the
structural dynamic motion control using double skin facades / distributed tuned mass
dampers.
Design integration among architecture-related disciplines is emphasized throughout the
research process as a means to more effectively overcome or at least minimize
contemporary technological limitations and to create architecture of higher quality.
While each study makes its own contribution theoretically and in a particular design
situation, from a wider viewpoint, the contribution of this thesis is to create more
constructive relationships of architecture-related disciplines to produce better architecture
through synergistic effects.
Acknowledgements
I am very grateful to Professor John E. Fernandez for his guidance and encouragement
throughout my research work. He has always been open-minded and encouraged me to
explore potentials and to pursue my interests. His enthusiasm as a teacher is inspiring,
and his wise suggestions have always been very helpful. This work would not have been
possible without his ceaseless support.
I also would like to express my deepest gratitude to Professor Jerome J. Connor for his
invaluable guidance and support. His capacity for clear scientific thinking led this work
to become a valuable intellectual contribution that will hopefully help the practice of tall
building design go in a more integrative and efficient direction. Working with him has
been one of the greatest experiences of my life at MIT.
Special thanks go to Professor Michael K. Kim for his continual support, confidence, and
Table of Contents
INTRODUCTION
19
23
27
28
35
38
48
53
53
55
56
3.4. Conclusion
58
63
63
65
66
71
71
71
76
79
79
81
81
83
87
89
9
90
91
91
93
95
98
101
103
103
6.7. Conclusion
104
105
105
107
107
111
113
113
114
114
119
126
127
131
131
134
135
138
139
140
143
143
147
150
153
153
CONCLUSION
155
10
APPENDICES
Appendix 1: Architectural/Structural Perception of Diagrid/Frame Structures
Appendix 2: Velocity Pressure for a 240m-tall 60-story Building in Boston
Appendix 3: LRFD Member Stress Check for Structures on Section 5.2
Appendix 4: Equivalent Plate Model of Diagrid Structures
Appendix 5: Preliminary Design of Diagrid Structures & Verification of Empirical
Equation 4.24: s = (
3 ,
159
161
165
167
169
187
191
Appendix 9: 2 DOF + 2 TMD H, Hdl & Hd2 for the Approximate Solution
209
Appendix 10: Tuning TMDs regarding Acceleration for Structures shown in 7.2.2
211
215
219
221
REFERENCES
225
11
12
Table of Figures
2.1: Home Insurance Building, Chicago
2.2: Demolition Scene of Home Insurance Building
2.3: Mills Building, New York
2.4: American Surety Building, New York
2.5: Section Drawing of Havemyer Building, New York
2.6: Havemyer Building, New York
2.7: Masonic Temple, Chicago
2.8: Reliance Building, Chicago
2.9: Height Race from Park Row Building (1899) to Empire State Building (1931)
2.10: Structural Section Drawing of Woolworth Building
2.11: Structural Section of Empire State Building
2.12: Woolworth Building
2.13: Facade Construction Scene of New York Municipal Building
2.14: Facade Construction Scene of Empire State Building
2.15: Chicago Tribune Tower Competition Entries by Hood & Howel, Saarinen, Gropius
and Hilbersheimer
2.16: Carson Pierie Scott Department Store, Chicago
2.17: Hallidie Building, San Francisco
2.18: Glass Tower Design by Mies van der Rohe
2.19: Structural Plan Drawing of Gulf Oil Building, Pittsburgh
2.20: World Trade Center, New York
2.21: John Hancock Center, Chicago
2.22: Sears Tower, Chicago
13
14
6.22: Preliminary Member Sizing for the 60-Story Diagrid Structure with Various s
6.23: Diagrid Steel Tonnage for 60-story Structures
6.24: Diagrid Steel Tonnage for 42-story Structures
6.25: Member Sizes for 60-Story Diagrid Structures based on 2 Different Maximum
Displacement Criteria
6.26: Member Sizes for 42-Story Diagrid Structures based on 2 Different Maximum
Displacement Criteria
6.27: Diagrid Construction in Swiss Re Building
7.1:
7.2:
7.3:
7.4:
7.5:
n
2 d
n
1st and 2 d
15
8.18: Dynamic Response of the 60-Story Structure without TMD: Modal Displacement
Profile - Real Part - for the Primary Structure; Modal Damping Ratio without
Feedback; Profile of the Maximum Nodal Relative Displacement - Primary
Structure; and Profile of the Maximum Nodal Total Acceleration/g - Primary
Structure
8.19: Dynamic Response of the 60-Story Structure with a TMD tuned to the
1 St
Mode
nd
Mode
nd
8.23: Dynamic Response of the 60-Story Structure with Vertically Distributed TMDs
tuned to the 1 t and 2 d Modes when Forcing Period = 1St Mode Frequency of the
Structure: Modal Damping Ratio without Feedback; Profile of the Maximum Nodal
16
List of Tables
2.1: Amount of Steel used per Square Foot of Floor Area for Empire State Building, John
Hancock Center, and Sears Tower
2.2: High-rise Building Construction Statistics by Region
5.1: Strategies to Reduce Wind-Induced Response of Tall Buildings
6.1: 3D and 2D Plan Views of a Diagrid Structure Model
6.2: Diagrid Angle vs. Maximum Horizontal Displacement for 60-Story Diagrid
Structures with Corner Columns
6.3: Diagrid Angle vs. Maximum Horizontal Displacement for 60-Story Diagrid
Structures without Corner Columns
6.4: Preliminary Member Sizing for a 60-Story Diagrid Structure
6.5: Preliminary Member Sizing for a 42-Story Diagrid Structure
6.6: Comparison between Proposed and Conventional Methods of Preliminary Design of
Diagrid Structures
6.7: Maximum Horizontal Displacement for 60-Story Diagrid Structures with Various s
Values
6.8: Maximum Horizontal Displacement for 42-Story Diagrid Structures with Various s
Values
8.1: Vertically Distributed TMD Tuning Properties
17
18
INTRODUCTION
19
primarily for static loads and damping for dynamic loads. Today's most popular tall
building structural systems are various tubular structures, core supported outrigger
systems with super columns, and more recently diagrid structures. The main
characteristic of these structural systems is that they have their major lateral load resisting
systems
at the perimeter
of the building
in order to maximize
structural depth.
investigating diagrid structures mainly with regard to static lateral loads. Since diagrids
are located at the perimeter of the building and generally there are only slanted structural
members without any conventional vertical columns, their architectural implications are
significant, especially with respect to facade system design. From this perspective, the
architectural potentials of diagrid structures are also investigated.
With regard to dynamic loads, new technologies are investigated through design
integration between the structural and facade systems of tall buildings. Special focus is
given to double skin facade (DSF) systems considering their increasing significance in
conventional system, its comparative advantages and disadvantages are investigated first
from the viewpoint of the structural behavior of a building that adopts this system, and
then its architectural implications from the viewpoint of its contribution to both esthetics
and functions.
The impact of technology is significant in tall buildings due to their very tallness.
Consequently technology tends to govern the design of tall buildings more than that of
other building types. This trend may conflict with architectural aspects of tall buildings.
Good design involves resolving this possible conflict. It depends on the capability of
architects and engineers to transform any present challenges into the potentiality of
enhanced design integration. This thesis is devoted to demonstrating this potential
through collaborative works between architects and engineers.
20
PART I:
TECHNOLOGY AND ARCHITECTURAL STYLE OF TALL BUILDINGS
21
22
Technology has very often been considered in terms of its relation to science and its
ability to apply critical scientific discoveries. However, the relationship between
technology and art has been investigated less intensely. Nevertheless, as Bruno (1995)
has noted, the nature of technology is quite similar to that of art in the sense that both
create something for humans unlike science, which pursues truth of nature for its own
sake. Architecture is one of the disciplines that encompass both technology and art. In
Purely formal, spatial, and structural analyses all appear to be inadequate tools in the
search for a precise delineation of this architectural style." Gothic architecture's
dematerialization for opening up the walls through structural innovation was primarily to
introduce more light - "symbol of the grace of God" - into the building for religious
purposes. In addition, its pursuing unprecedented height through the structural invention
was for religious reasons. Thus, the fundamental driving force of the emergence of
Gothic architecture was not structural but absolutely religious, and this religious purpose
23
the essential nature of architecture of a certain age and eventually originating a new style.
Observing the history of technology, an epochal leap of technology rarely occurs. In
general, development of technology is not revolutionary but evolutionary. Mostly in very
rare revolutionary cases, new technology changes the nature of artifacts in terms of both
function/performance and esthetics. This phenomenon is in a sense natural because an
epoch-making technological leap means that there have been no precedents. In a case
when new technology creates an artifact of new configuration using new material to hold
new function/performance, if the artifact is in a discipline that encompasses art, new
esthetic expression is sought continuously and finally generalized into a new style. In
that case, this new technology may be called original technology in the sense that it
modifies the essential nature of an artifact by not only generating an unprecedented
function/performance but also originating a new style in terms of esthetics. Original
intrinsic nature of any technology is ephemeral and ever-evolving towards efficiency and
better performance. Thus, original technology is followed by refinements based on the
results of its performance. However, this new set of refining technology following the
original one does not change the nature of artifacts but evolves from the original one and
remedies its unforeseen problematic aspects to make it perform as intended or even better.
In this sense, this evolutionary technology following the revolutionary original one may
be called remedial technology and it is mostly in the function/performance domain.
technologies
technologies
are
much rarer than function/performance-driven technologies due to the fundamental nature of any
technologies.
24
away from traditional to modern. However, the other side of this promise was the
imperfection of experimental original technology and intrinsic ephemeral nature of any
technology. In addition, the power of technology superficially perceived by architects in
the early modern period was much stronger than the actual capability of technology.
Modern architecture was born in this physical and mental gap. This circumstance led to
the next step of technology that smoothly fills and connects this gap in order to augment
the integration between technology and architecture originally produced by technological
revolution. This is remedial technology, distinguished from millennia-breaking original
technology. Modernism, as Steele (1997) noted, despite repeated declarations to the
contrary, is far from dead and may now be entering its culminating phase through
remedial technology.
25
26
Various technologies have been involved in the design and construction of modern
buildings. Also, since the emergence of modern architecture many stylistic transitions
have occurred. Some have been global, and others have been regional. Some have lasted
longer, and others lasted a shorter time. Technological evolutions and technology itself
are fundamentally in the domain of function/performance
However, this chapter hypothesizes that some technologies have had major impacts on
the esthetics of architecture as well due to the intrinsic nature of architecture, which must
satisfy not only functional but also esthetic requirements. Consequently, some other
technologies have had only minor impacts, and the rest of the technologies have had no
impact at all on the esthetics of architecture.
Tall buildings are, in a sense, the accumulation of the most advanced modern
architectural technologies due to their extreme height. Because of their enormous scale,
which comes again from their height, the impacts of their architectural esthetic
expressions are also significant in any context where they soar. Considering this special
significance of the technologies and esthetics in tall buildings, the second hypothesis in
this chapter is that the interactions between these two issues have been more conspicuous
in tall buildings than in any other building type. Based on both hypotheses, the purpose
of this chapter is to investigate the interrelationship between the technological evolution
and architectural esthetics - further architectural styles - of tall buildings from their
emergence in the late 19 th century to the present. Among the various technologies
involved, structural systems and facade systems are mainly investigated in this thesis.
They are considered to be the most important technologies in tall buildings in terms of
not only their fundamental function/performance, which allows for the very existence of
tall buildings, but also their major influence on architectural esthetics and, in turn, style.
Finding symbiotic interrelationships between technology and esthetics is fraught with
misleading and sometimes erroneous linkages.
Tracing both chronologically,
technologies follow evolutionary paths with fluctuating rates; and styles, in a sense,
follow cyclic paths with varying periods. From a wider viewpoint, this chapter
investigates the way of their marriages and divorces following their historical paths up to
the present and the reasons for the paths. From a narrower viewpoint, this chapter studies
specifically which technologies are in the domain of both function/performance and
esthetics, which other technologies are in the domain of only function/performance, and
what the specific characteristics of these technologies of different natures are. In this
way, through the filter of historical facts, current situations regarding these relationships
can be clearly illuminated. Further, based on that, more desirable visions for the future
may be constructed.
The investigation of the relationships is conducted through the case studies of tall
buildings, many of which are still standing and some demolished for various reasons,
mostly economical ones. The choices of buildings studied are based on the significance
of applied technologies and that of esthetic expression. Thus, considering the nature of
the field of architecture, the buildings chosen for esthetic reasons are all renowned ones
27
Governs,
Technology
~j--= --
For example, the ceiling heights of the Amiens Cathedral and the Beauvais Cathedral are 138 feet and 158
feet respectively.
The heights of the Western Union Building and Tribune Tower were 230 feet and 260 feet respectively.
The heights of these multi-story office buildings were possible due to the invention of the automated
vertical transportation system - elevators - which is one of the most important technologies for the
emergence of tall buildings. However, the discussion on elevators is limited here because it is less relevant
to the theme of this thesis. The first passenger elevator, designed by Elisha Otis, was installed in the 5-story
Haughwout Building in 1857. For office buildings, the first use of elevators was in the Equitable Building
in New York in 1870.
28
the traditional load-bearing masonry structures, which required the wall thicknesses of
several feet on their ground levels.4 Thus, these earlier tall office buildings did not have
the potential to grow further because of the technological limitations of their structural
system. In these technological contexts in both New York and Chicago - the only two
skyscraper cities in the world at that time - the invention of the iron/steel skeletal
structure for the Home Insurance Building was a remarkable breakthrough toward the
development of new building type.
In fact, the invention of the iron/steel skeletal structure was driven not by the structural
aspiration to build buildings higher but rather by its economic purpose to introduce as
much natural daylight as possible into the interior office spaces to attract higher rents.
This very well explains the fundamental nature of technology as a means to an end in the
"As it was important in the Home Insurance Building to obtain a large number of small
offices with abundance of light, the piers between the windows were reduced to the
minimum, and thefollowing system of constructionwas adopted. Iron was used as the
skeleton of the entire building except the party walls, and every piece of iron was
protected from fire by masonry, excepting only some columns so situated as not to be
dangerous if left exposed. ,6
With regard to the technological achievement of the Home Insurance Building,
architectural historian Carl Condit (1968) was skeptical about the general acceptance of
this building as the first skyscraper because there were no provisions for lateral wind
4 For instance, the wall thicknesses of the Tribune Building on the ground floor varied from 3 feet 4 inches
to 5 feet 2 inches.
5Generally,
the thickness of the brick masonry wall at the topmost story of the building was 12 inches with
4 inch increments every story downward to the ground. For example, the wall thickness on the ground of
In his 1934 Architectural Record article Neither a Skyscraper nor of Skeleton Construction, architect
Irving Pond was opposed to calling the Home Insurance Building the first skyscraper due to masonry party
walls. However, it was inevitable because Chicago's building code at that time did not accept iron/steel
6
29
loads in this 10-story building. Actually, there is no generally accepted definition of tall
buildings. However it is widely accepted that, from a structural viewpoint, if the tallness
of a building significantly influences its structural design and performance, that building
is considered a tall building. In fact, the tallness of the Home Insurance Building had
little effect on its structural design, resulting in Condit's skepticism. However, the
importance of any technological breakthrough should be measured not by how close it is
to the fully developed conditions but by its potential for further development. In the case
of the steel skeletal structure, the Chicago invention in the late 19t h century eventually led
And there lies its significance. Since the
to the present status of tall buildings.
construction of the Home Insurance Building, its iron/steel skeletal structure has become
the major prototype for the structural systems of tall buildings.
Following the emergence of iron/steel skeletal frame structure, various lateral load
resisting systems were developed. Systems developed in the late l9 t h century were
riveted steel connections, portal bracings, and braced frames. Riveted connections were
introduced in Holabird and Roche's Tacoma Building (1889) in Chicago. Portal bracings
were employed first in Burnham and Root's Monadnock Building (1891) and Jenny's
Manhattan Building (1891) in Chicago. And braced frames were used widely. These
series of structural innovations, occurring within the real estate boom in the late 1880s in
Chicago, established a solid technological foundation for much taller buildings to come.
However, following the financial panic in 1893 due to the excessive development of tall
buildings in previous years, the Chicago City Council was forced to limit the height of
buildings to 130 feet. This height limitation over the next three decades made the
development of tall buildings in Chicago much less active than in New York (Carol
Willis, 1995).7
With the invention of steel skeletal structures, the curtainwall concept was consequently
developed to clad them. The biggest challenge for architects confronted with an
unprecedented building type and new technology was how to design the fagades of tall
buildings, especially in terms of esthetics. What should be clearly noted here is that
esthetic expression was not within the initial criteria of the development of new
technology. However, due to the nature of architecture, new esthetic expression
corresponding
Strategies differed in
New York and Chicago, and various factors influenced the design of tall buildings in the
two skyscraper cities.
New York is a much older city than Chicago, and thus there was a prevalent architectural
stylistic trend already established. Naturally, this existing trend was directly applied to
early masonry load-bearing tall buildings. Thus, most early tall buildings looked odd
because existing styles were developed not for a new building type, tall buildings, but for
buildings usually lower than five stories. This oddness came from the awkward vertical
grouping of tall buildings. The early skyscraper architects had a hard time finding out
how to deal with this unprecedented tallness. Thus, they often located vertical grouping
design devices, such as cornices, at every two or three stories, or sometimes irregularly,
7 The height limitation fluctuated
from minimum 130 feet (1893) to maximum 264 feet (1921) over the
next three decades before the first Zoning Ordinance of Chicago in 1923.
30
"It is true that the skyscraper is in fact a series of equal cells, and
that the only suggestions for a triple division that inhere in the
conditions are the facts that the ground .floor has a d(fferent
destination from that of the floors above, and suggest a distinctive
treatment of the bottom, and the fact that a visible roof or in
default of it the necessity for a protective and projecting cornice,
compels a distinctive treatment for the top.
Most successful of the skyscrapers are those in which the shaft is
made nothing of in which necessary openings occur at the
necessary places, are justified by their necessity but draw no
attention to themselves. "
At the very end of the century, this comment about pure
functionalism is in exact agreement with Sullivan's dictum,
"Form Follows Function," 9 and also Mies's "Less is More"
decades later.
He accurately predicted the 20th century
architecture. In the same article, he clearly stated the inadequacy
of applying masonry to the steel skeleton as its fac;ade material.
Figure 2.4: American
Surety Building
Even though the tripartite fa<;ade design approach prevailed at the end of the 19th century, the previous
vertical grouping design was still applied for some buildings such as the Park Row Building completed in
1889, the tallest building at the time of its completion.
Sullivan said that the "lower two (or possibly three) stories of a tall office building had a destination so
different from that of the superstructure, that a distinguishing treatment for them was not only required but
demanded, and that the uppermost story in turn, being in great part devoted to the circulation system of the
building should also be differentiated." Even though he declared "form ever follows function," in reality,
his form sometimes followed his esthetic viewpoint.
As Hoffman (1998) pointed out, in his early
masterpiece, Wainwright Building, the second floor and typical floor plan are exactly identical offices.
However, the second floor was not grouped with typical floors to compose the shaft but grouped with the
first floor to comprise the base. This design decision came clearly not from functional but from esthetic
considerations.
31
II For instance, John Moser's 1894 article in "American Architect and Building News" reads as follows.
"to express, not hide or falsify, our incomparable modern steel construction, .... I found myself getting very
close to the Classic."
32
The fact that Chicago has a much shorter history than New York, and the Great Fire in
1871 burned down almost the entire city, made Chicago's design of tall buildings
different from New York designs. There was no existing urban context to follow in
Chicago at the early stage of skyscraper development. Naturally this allowed Chicago
architects to be free from existing styles. This is one of the most important factors of the
so-called "Commercial Style" of the first Chicago School. 12
The Chicago School's design approach to tall buildings can be observed well from John
Root's 1890 remark about modern business buildings.
"...to lavish upon them profusion of delicate ornament is worse than useless... Rather
should they by their mass and proportion convey in some large elemental sense an idea
of the great, stable, conservingforces of moderncivilization."
The idea behind this remark is the same as that of Adolf Loos, the author of Ornament
and Crime in the early 1900s. The same idea was manifested decades later by Henry-
12The influence of William LeBaron Jenny, the inventor of iron/steel skeletal frame for tall buildings, can
be considered as another very important factor. Jenny was an architect/engineer
an engineer in France. Naturally he designed his tall buildings using new technologies and applying the
simplest forms without much ornament compared with his New York counterparts at that time. Most
important skyscraper architects such as Louis Sullivan, Martin Roche, Holabird, and Daniel Burnham were
trained at his firm and it is very probable that Jenny's approach directly or indirectly influenced their later
tall building designs.
'3 The Home Insurance Building has an arched facade also at its second level, which is actually masonry
structure. (The first and second levels of this building are masonry structure. The upper levels from the
33
15 While the upper four stories of the old existing building on the site were held up on screws due to the
remaining lease, the construction of the foundation and the first story of the new Reliance Building was
performed.
34
2.2. Skyscrapers in the Early 20th Century: Return to Schism in the U.S. vs. the
Modern Movement in European
As was discussed in the rrevious section, from
around the end of the 19t 1 century Chicago was
not as active as New York in developing tall
buildings.
Despite the fact that Chicago
accomplished
significant
technological
development for further growth, capping the
height due to the financial panic in 1893 made the
development of tall buildings shrink. Thus, from
that point, New York became the leading city of
tall buildings in terms of both amount of
. an dh elg
. h ts. 16
constructlon
00
I!
J
0
a:
Co
::.
S a
II.
\:r)
!
U
:5
c
i"
Q
....
;,
0
i
lit:
15
~
~
Co,)
fi
I
~
:;
i...
~
~"
i !
IE
Even though the heights of skyscrapers were significantly increased during this period,
contrary
intuition,systems,
there had
much conspicuous
technological evolution. In
terms
of to
structural
mostnottallbeen
buildings
in the early 20
th
century employed steel
braced frames just as those built during the previous century did. Among them are the
renowned Woolworth Building and the Empire State Building completed in 1913 and
1931 respectively.
Their enormous heights at that time were accomplished not through
notable technological evolution but through excessive use of structural materials. Due to
the absence of advanced structural analysis techniques, they were quite over-designed. It
can be noticed from the amount of steel used per square foot of floor area (Table 2.1) and
16 According to Carol Willis (1995), as late as 1923, only 92 buildings of 10 to 22 stories were built in the
Loop, while almost 1000 buildings of 11 to 20 stories were built in Manhattan by 1913. "During the boom
of the 1920s, New York, with its more active speculative environment, pulled farther ahead of Chicago in
the number and height of its towers. The tallest building in the Loop in 1931 was the 45-story, 612 ft
Board of Trade; in Manhattan, 16 spires exceeded that height, including the Empire State Building, at 1250
ft."
35
37
37
However, as can be seen from the various international entries for the Chicago Tribune
Tower Competition in 1922 (Figure 2.15), the rebirth of the early Chicago spirit and the
application of European modem movements in tall buildings were only a matter of time.
Even though Hood and Howell's Gothic style entry won the competition, many entries
such as ones by Gropius and Hilbersheimer clearly denoted the upcoming International
Style skyscrapers.
Saarinen
Gropius
Hilbersheimer
Style / Modernism
"Each new style gradually emerged from the earlier one when new methods of
construction, new materials, new human tasks and viewpoints demanded a change or
reconstitution of existing forms." - Otto Wagner In 1828, German architect Heinrich Hubsch posed a serious architectural question, "In
What Style Should We Build?" With regard to architectural style, the 19th century was a
transitional period. In the absence of a strongly governing architectural style, architects
tried to find stylistic solutions from Classical or Medieval ones. Instead of stone - the
major material for traditional architecture and civil structures - the new material, iron,
had begun to be used in civil structures such as bridges and incidentally in architecture
from the late 18th century. However, at the time of Hubsch's publication, the use of iron
in architecture was not significant. Some embryo of new style began to grow, but it was
not conspicuous until around the turn of the century.
A stylistic breakthrough using iron and glass as building materials can be found from the
renowned Crystal Palace constructed during the mid-19th century. The Crystal Palace is,
as Nikolaus Pevsner noted, "the mid-nineteenth century touchstone, if one wishes to
38
discover what belongs wholly to the nineteenth century and what points forward into the
twentieth." This was the skeletal structure with glass skin, "architecture as volume" with
"regularity" without "applied decorations." It literally fitted the definition of the
International Style of the mid-20t h century. However, architects and critics did not
welcome this new architecture at the time of its construction, and, as a consequence, it
did not immediately further grow as a new style. For instance, Pugin called the Crystal
Palace a "glass monster."
The most significant use of iron/steel and glass in architecture emerged during the late
19th century in the U.S. The development of a new building type - skyscrapers necessitated the use of iron - later steel - skeleton and glass skin. The iron/steel skeletal
structure for tall buildings was the invention of the Chicago architects as was discussed
earlier. 8 In terms of style, early tall buildings in Chicago foresaw the emergence of the
International Style decades later.
In his book, Hubsch noted that climate and building material are the two principal
formative factors of style. They are fundamental considerations in any architecture,
which must be built with proper materials on a particular site of a certain climate. Thus,
in describing each building element that is the very component of any specific
architectural style, his statement regarding climatic condition is quite elaborate.
"In our northern climate, more care must be taken to protect buildings against rain or
snow than in the south. The pitch of the roof, normally covered with slate or perhaps
even with tiles, must be steeper than on Greek monuments. Furthermore, the projecting
upper surface of cornices or other similarparts must have a distinct slopefor the water
to run off .... If the surface of a projecting cornice is not sufficiently sloped, this surface
itself will suffer from weathering. Decay will soon set in .... "
About a century later, the definition of the International Style by Hitchcock and Johnson
was concerned with only esthetic qualities of the style as Barr noted. Comparing the
description of the roof by Hubsch with that by Hitchcock and Johnson gives interesting
insight.
.... the flat roofs normal with modern methods of construction have an essential aesthetic
significance. Roofs with a single slant, however, have occasionally been used with
success. For they are less massive and simpler than the gabled roofs usual on the
buildings of the past. Flat roofs are so much more useful that slanting or roundedroofs
are only exceptionallyjustified
Climatic considerations could not be found in their book, The International Style. This
ignorance of the importance of "uncommon ground" in architecture was based on a too
18 It
is interesting to note that New York had much more potential to initiate tall buildings than Chicago
with Bogardus' works of iron frames that contained essential features of skeletal frames and curtainwall
concepts. However, ironically as Condit pointed out tall buildings emerged from Chicago where iron
works were not as developed as New York.
39
The mid-twentieth century, after the war, was the era of mass
production based on the style defined, and technologies
developed earlier. The major driving force of tall building
developments was economy. Even the once-prevalent height
race did not occur after the war until the construction of the
World Trade Center and the Sears Tower, completed in 1973
and 1974, respectively.
Most tall buildings of the midtwentieth century employed basically the same structural
systenls, steel skeletal frames with diagonal wind bracings,
developed in the previous century. As the forms of many tall
buildings were pure rectangular boxes, even some structural
Figure 2.18: Glass Tower
19 Sullivan commented
about the World Faire that "the damage wrought to this country by the Chicago
World Faire will last a half a century." And as Gideon noted, his prediction was generally what happened
to architecture in the U.S.
40
20
41
There also have been some other structural approaches for tall buildings such as
staggered truss system developed by William LeMessurier at MIT in the 1970s and
stressed steel skins employed for the One Mellon Bank Center in the 1980s. All of these
structural variations, coming from the efforts to find more efficient systems than alreadytraditional steel braced frames to satisfy varying requirements, are on the general
evolutionary track of technology. Primarily, they are driven by function/performance.
However, depending on the configuration of each system, some systems such as tubular
structures and diagrid structures have major impact on the building esthetics as well. In
addition, some tall buildings, such as Hotel de las Artas in Barcelona (Figure 2.25) or
42
21 .. We
mostly live in enclosed rooms. These form the environment that gives rise to our culture. In a way
our culture is a product of our architecture. In order to raise our culture to a higher level, we are forced,
whether we like it or not, to change our architecture. And this will be possible only ifwefree the rooms in
which we live of their enclosed character.
This, however, we can only do by introducing a glass
architecture, which admits the light of the sun, of the moon, and of the stars into the rooms, not only
through a few windows, but through as many walls as feasible ...... The new environment we produce in
this way has to bring us a new culture." - Sheerbart In fact, by comparing Gropius and Corbusier's works, Rowe and Slutzky evaluated Corbusier's
phenomenal transparency as having much superior esthetic value than Gropius' literal transparency, which
was achieved by actual transparent glasses.
22
43
Frederick
Koeper
well describes
the scale-wise
characteristic of the decoration applied to the exterior of
modernistic tall buildings.
"Curiously and not altogether appropriately, this style of
surface ornament and rich materials, so effective in
interior design and objects of luxury, was applied to the
outside of giant skyscrapers.
The small scale
characteristic of Art Deco was hardly suited to the bulk
and economic purpose of the skyscraper; yet Art Deco
ornament satisfied a certain self-conscious urban taste for
modernity, albeit much of it is lost to the distant eye. ,,23
Tall buildings employin~ truly modern style facades began
to be built after the war. 4 A notable example is the SOM-
Built in the transitional period, the McGraw Building and Daily News Building stand stylistically
somewhere between Modern and Art Deco.
23
44
designed Lever House (Figure 2.29). The original curtainwall of the Lever House
constructed a half century ago was con1posed of stainless steel mullions (supported fron1
inside by carbon steel channels and angles), blue-green wire-glass spandrel, and light
green tinted vision glass (Figure 2.30). With its very thin mullions and clear light tinted
glass, it has the quality of "literal transparency" and through this transparency the
structural columns inside are visible. Architecture as a volume with regularity without
applied ornaments was truly achieved by modern technology in the Lever House. New
architectural style incorporating new technology thus reflecting the spirit of the age was
accomplished in the Lever House and numerous ilnitation works followed after this
building.
EXPlOOlD
!vnt;lure of
DIAORAM.
fiorizonlol
mvlli.on and
vertical
reil wl,jci, guid~s window'wo~I,i"9
gondola .
Time did not flow without giving any penalty, however, for using the technology in its
infancy. Most original glass curtainwall systems on the fa<;ade of the Lever House failed,
and as a consequence this building recently went through an entire exterior curtainwall
restoration?5 Due to its significant position in the history of modern architecture, debates
horizontal bands of aluminum windows and brick spandrels. Also, PSFS was the second major building
equipped with modem mechanical HV AC systems. This more than 70 year old building is still well
functioning without major restoration works, especially with regard to fayade, which has much to do with
the style.
25 Even though
a drainage mechanism was provided in the mullion system, due to the inadequate
characteristics of polysulfide as a sealant and also inadequate provision of weep holes, water penetrated
through the stainless steel mullion cover, corroding the carbon channels and the ends of wire in the wireglass spandrels. As the wire expanded due to the corrosion and also because it is inherently less resistant to
45
on authenticity prevailed at the time of the restoration of the fa~ades of the Lever House
and it left a question regarding one extreme branch of metaphors of modem architecture,
the metaphor of the machine initially conceived by Viollet-Ie-Duc. If considered as a
machine, any components of a modem building can be replaced whenever their longevity
is reached without considering authenticity because they were designed as parts of a
machine to make it fully functional. Thus, if a copy of the original part is not available
due to the time gap, any compatible part that can make the machine fully functional or
possibly even more functional can replace the original part.
The restoration process of the Lever House was, however, different from that of any
machine part. To preserve the Lever House built a half century ago, exact imitation was
deliberately pursued. Light green tinted single pane vision glass panels, which represent
the status of the technology at the time of original construction but do not correspond to
the technology and energy-conscious design at the time of restoration, were selected in
order to give the building fa~ade exactly the same tone and transparency as the original.
Moreover, fake mullions were introduced at the middle of the spandrel panels (Figure
2.32). The original spandrel panels were composed of two panes of blue-green wireglasses connected by mullions in-between, due to the limitation of size available at the
time of original construction (Figure 2.31). However, the blue-green back-painted
spandrel glass panels used for restoration works were single panes divided by nonfunctioning fake mullions at the mid-height in order to imitate the original look.
['(\$tInt .~tkal
to n:!melfl
g11m
': :
,top
III
Continuous
~ellass
Edlt.nll sWn!e<J$.Iteet
--CO':'lr5 to br. rel'l'l(:lVe(l
an<: rep!8Ce:l
..
..
1.1
d\.l{ii
temperature fluctuation, the spandrel panels began to crack, leaving almost no original panels at the time of
the restoration. Some vision glass panels also cracked, possibly due to the compression from the hardened
polysulfide sealant, thermal movement, and corrosion of carbon steel channels.
46
47
See-through Miesian glass boxes with hierachical articulations of structures and mullions
transformed into geometric forms of mirrored glasses that reflect the surroundings in the
daytime and expose Miesian articulations at night. Recently, point fixings have been used
for glass favade constructions for mullionless surfaces, producing somewhat different
esthetics from what structural silicones generate.
More recent environmental design strategy has produced many notable tall buildings that
employ various advanced technologies such as double skin facades.26 They are on the
general path of performance-based technological evolution. However, their esthetic
ilnpact is not ignorable. Their perceptual visual transparency is much greater than
conventional single skin facades. (Thus, not surprisingly in the field of architecture,
some architects use this system primarily for esthetic reasons and justify it with its better
performance.)
This design strategy is in the continuing mainstream of modem
architecture today.
Stiffening
trusses
Steel plates
Rigid frame
sleelluoo
TranSfer
from
lub\) \0 sloo\
shear boxes
Steel plate
shear boxes
Figure 2.35:
AT&T
Ruildim!
Double skin facades showed up intermittently in the U.S. such as Warren Petroleum Building in Tulsa in
1950s and Occidental Chemical Center in 1980s.
26
48
Behind these traditional gestures were apparently modern technologies - steel skeletal
structures and curtainwall concept a century old already. Thus, post-modern tall
buildings can be interpreted as the products of reaction to the evolutionary pace of
modern technology compared with the revolutionary one of the previous century. They
camouflage a hundred-year-old technological solution with traditional-looking fa9ades
produced and hung by modern technology. Indeed, with the lapse of time accompanying
technological advances, architectural styles have accumulated and it has become possible
for architects to choose any style from a broader library. However, the choice should be
made with zeitgeist. Arbitrary choices may create only an anachronistic architectural
environment.
Another type of reaction to the evolutionary pace of technology is ironically exaggerated
advocatative expression of technology through the design of building exteriors. In terms
of actually applied technology, so-called high-tech architecture still employs general
contemporary technology in most cases. However, the architecture of this trend tries to
elaborately express it in order to pretend that there has been a significant technological
evolution. Even though technological evolution generally accompanies economy of
building construction, the architecture of this style is never economic. In fact, it is much
more expensive.
Comparison of the statistics of tall building developments between overall and the last
ten years denotes that the most active tall building development region has been shifting
from North America to Asia over the last decade (Table 2.2 from www.skyscrapers.com).
In terms of architectural design, the most significant trend of those tall buildings
constructed in various Asian countries is that they use their own regional architectural
traditions as main design motives. This trend can be easily seen from notable recent tall
buildings such as the Jin Mao Building in Shanghai (Figure 2.36), Petronas Tower in
Kuala Lumpur (Figure 2.37), and Landmark Tower in Yokohama (Figure 2. 38).
HIGH-RISE BUILDINGS BY REGION (OVERALL)
Region
Bldgs
Percent
17,765
51.49%
Region
Bldgs
Percent
Asia
1,797
32.67%
1,711
31.11%
N. America
Europe
5,881
1704%/
N. America
Asia
5,412
15.69%
Europe
905S
16,45%~
Middle East
2,071
6.00%
Middle East
488
8.87%
Oceania
1,637
4.74%
Oceania
335
6.09%
S. America
1,352
3.92%
S. America
213
3.87%
Africa
386
1.12%
Africa
51
0.93%
.-
Table 2.2
49
The history of Regionalism has comes with the history of architecture from Roman
Vitruvius to recent Critical Regionalism of Tzonis, Lefaivre and Frampton. Regional
architecture was understood very skeptically by many architects of the International Style.
This can be well observed from the "What is happening in modem architecture?"
conference held in the MoMA in 1948. Many important figures in modem architecture
such as Barr, Hitchcock, Gropius, Blake, Mumford debated on International Style and
Regionalism without any fruitful results. Below is part of Mumford's presentation.
Certainly his remark did not receive any agreement from architects of the International
Style at that time.
"Regionalism has to help people come to grips with the actual conditions of life that
make them feel at home. Regional insight has to be used to defend us from the
international style, the absurdities of present technology and the despotism of the
mechanical order. "
Contemporary tall buildings in Asian countries may be categorized as romantic
regionalism architecture. These buildings pursue only the images of the traditional
27
architecture of their regions, and their designers are in many cases American architects.
This is the marriage of the image of a particular region and the technology of the
Jin Mao building was designed by Skidmore Owings and Merrill, Petronas Tower by Cesar Pelli, and
Landmark Tower by Stab ins Associates.
27
50
International Style.
architecture in general.
Another reactionary architectural trend has been so-called deconstructivism since the
early 1980s. While post-modern architecture mainly reacts to form and surface of the
International Style, deconstructivist architecture does not consider rational planning
necessary. Even though there have not been any notable tall buildings built in
deconstructivism, some proposals were made by Peter Eisenman and Frank Gehry.28
Frank Gehry's works were included in Deconstructivist Architecture exhibition held in the MoMA.
However, Gehry does not consider himself as a deconstructist architect.
28
51
52
53
and glass curtainwall systems clearly anticipated new stylistic implications, just as ribbed
vaults and flying buttresses had generated Gothic architecture centuries ago. As Turner
(1986) pointed out, "a time lag exists between the introduction of new technology and its
large scale application to major building projects and a further chronological gap appears
between the technological advance and stylistic incorporation of it." However, observing
the evolution of tali buildings, the time gaps between these events were very short. In
fact, in Chicago, the invention of iron/steel skeletal structures, the application of it to
lnajor building projects, and its stylistic incorporation occurred almost sinlultaneously in
the late 19th century, generating the first Chicago School's "Commercial Style.,,30
Beginning in the early 1930s and growing strongly after the war, the International Style
became the new culmination of technology and architecture. Arising from the stylistic
elnbryo of steel framing and glass curtainwalls introduced in Chicago, the International
Style was born (Figure 3.1 and Figure 3.2). The new technology of lightness and
transparency fully incorporated the new esthetics, which so consciously expressed the
ideology of current technology in terms of its functional goals and performance attributes.
In this sense, the new technology of modern steel framing and curtainwall assemblies
produced a new building type for the new function of the modern office space. This
generated a new esthetic expression, and ultimately originated a new architectural style.
For this reason the new technologies of braced steel skeletal structures and glass
This stylistic transition, which began at the turn of the century, was interrupted for about three decades.
During the early twentieth century, while European architects were going toward a new reductionistic
movement, American counterparts involved in tall building design indulged in eclectic styles, especially
Art Deco, from the 1920s to the early 1930s. This phenomenon might be understood as an effort of
architects to recover the representational characteristics of architecture lost due to the new technology.
Thus, they tried to return to the traditional architecture for representation ironically with new technologies.
However, as was discussed in Section 2.2 of this thesis, application of modern movements in tall buildings
was only a matter of time. For example, even though Hood and Howell's Gothic style entry won the
Chicago Tribune Tower competition in 1922, many entries, such as one by Gropius and Hilbersheimer,
clearly denoted upcoming International Style skyscrapers.
30
54
curtainwalls may be called "original technology." The 1950s and 1960s were the era of
production based on this original technology, without notable technological evolution.
As the technologies of these buildings developed and the interest in the esthetics of the
International Style grew, the tall buildings of that period were becoming literally
"international". Miesian tall buildings could be found not only in Chicago and New York
but also in major cities all over the world.
internationally and the esthetic interests that supported and resulted from these new
assemblies were also adopted globally.
Thus, even though the tall buildings that employ the tubular concept as their structural
systems are quite expressive due to the configuration of the system, it did not shift the
intrinsic nature of tall buildings in terms of function and esthetics.
In fact, as a result of increasingly economical structural solutions using less material
arranged more efficiently, a new structural motion problem was created. Tall buildings
were now safe but not stiff enough for human comfort, a problem not anticipated during
the earlier days of structural innovation for these kinds of buildings. For example, the
Empire State Building, completed in 1931, could have been designed more efficiently,
using less material, if engineered today. The lack of structural analysis tools meant that
those involved in its design were understandably conservative in the amount of material
necessary for a safe structure. However, this building did not produce any motioninduced human discomfort problem despite its unprecedented enormous height. It was
structurally over-designed by today's standards and thus stiff enough not to cause that
problem. In contrast, even though 1960s and 1970s' tall buildings, which adopted much
lighter structural systems, achieved higher efficiency, they caused serious motioninduced human discomfort problems. This comparison demonstrates the imperfect
character of the introduction of original technologies and the necessity for ameliorating
steps in subsequent development.
And yet, it is clear that the evolution of structural systems for tall buildings in the
address the issue of excessive motion in towers, do not change the nature of tall buildings
55
!? lJ \\\
~ -(7n t!!
TORSIOflAl SWAY
300 TON
MASS BlOCK
222FT
343FT
There were some opposing trends with regard to transparency intervening between the emergence of tall
th
buildings in the late 19th ce~tury and the culmination of them in the mid 20 century after World War II.
For instance, renowned Art Deco skyscraper architect Ralf Walker's concept of architectural space was
opposed to that of transparent volumes. He notes that "to merely build a shed with one or two walls of
glass does not create space ... it merely interrupts it." While American eclectic architects in the early 20th
century were going in a direction somewhat opposed to transparency, their European counterparts pursued
it more rigorously. Notable examples among them are Glass Chain's pursuit for transparency through glass
in architecture, Mies' glass skyscraper projects during the 1920s, Gropius' Bauhaus design in Dessau, etc.
31
In fact, fluorescent lamps introduced commercially in April 1938 and displayed publicly in 1939 at the
New York World's Fair and the Golden Gate Exposition in San Francisco made the role of natural light
much less important than earlier days. According to Murdoch, from 1952, fluorescent lamps passed
incandescent lamps as the major source of general lighting in the United States.
32
56
, ~~ 01[.Q i~ de
tmh<!t ,r;j
and
lit
l(Il) ~
";,,alI4:1!
I
CcltntcUl!. 'm~roo~ to
:tIu'a- Y.l.1icn ,~jl'l';~ ;0 nrt<
~l:l':!lbr,~i1:tt'"
Ii
, <VIj.l:J'llrret.'~OVir
tNrre .,~
~ 1inI~ lop. TheCCIi"i tMllll~
~_~ COITlJl!ssi:ln
!Il:)t
of Ih f~t:1 ~ r~
t; C't-
,-/1 .. It:f
etl1l
m\".C
fi'I!<l\\1thgr()l,t~~~n
1ll.~.!Jb"lt!lY3;IC!,ll.
, SI/fljcSs20'iY..
I
, Ftm"
1~
roc:~ C!!*'Hhlt
10
Figure 3.5: Metal Decking Used for Exterior Walls for Ventilation in Beard House
In fact, maintaining the role of the transparent glass fa9ade as an environmental mediator
was not neglected at the beginning of the modern movement, especially in Europe. The
year 1903 already saw the very early example of glass double skin fa9ade in the Steiff
Factory building in Germany. In addition, Paul Scheerbart, while declaring glass as the
only proper material for new architecture that can raise our culture to a higher level,
proposed the double glass wall as an essential condition for glass architecture,
recognizing the inadequacy of the single pane glass wall as an environmental mediator
and the role of air as a very effective insulation medium.33 Also, Le Corbusier's mur
neutralisant concept, inspired by Maison Isotherme by Raoul Decourt (Figure 3.5),
employed a double layer exterior wall, within which conditioned air circulated. In the
U.S., even though not a glass wall in particular, Neutra's Beard House in 1935 used metal
decking for its exterior wall in order to achieve fa9ade ventilation to provide a more
comfortable interior environment (Figure 3.6).
From the early 1950s, the double layer glass, Thermopane, was commercially available.
Also, in Germany in 1962, the thermal break was introduced for curtainwall mullion
systems. However, it was not until the early 1970s' OPEC oil embargo that energy
consumption by buildings was more seriously considered than ever. Recognizing the
problem of single pane transparent glass fa9ades, numerous remedial technologies,
including reflective glass, insulating glasses, low-E coatings, double skin facades, etc.,
have been developed. Among them, double skin facades might be considered as the most
advanced technology from the viewpoint of performance. Double skin facades being
remedial technology, most tall buildings employing glass double skin facades as an
exterior envelope system stylistically fall within the category of Modernism in this
esthetically plural age. Perceptually, double skin facades produce even more transparent
expreSSIon.
33 "As air is one of the worst conductors of heat, the double glass wall is an essential condition for all
glass architecture. The walls can be one meter apart - or even farther ... " - Sheerbart-
57
Today, tall buildings with glass double skin facades can be easily found in Europe. As
was discussed earlier, the environmental consideration of workplaces was somehow lost
in early days, especially after the Hawthorne Effect in 1924. And ineffective modern
glass fa<;ade was managed by consumption of more energy. However, consideration of
work place environments for better productivity began to be reconsidered by architects
and environmental psychologists from the 1960s. Since then, while Europe has been
continuously pursuing this goal toward its limit, the interest of the American counterparts
has not been as strong as that of Europe. This explains the more advanced status of
fa<;adetechnology in Europe. In the U.S. no notable tall buildings have employed double
skin facades. It can be understood in the way that Europe has been much more sensitive
to environmental design and also has a much longer history of double skin facades than
the U.S. Earlier double skin facades can be found in the U.S. in the Warren Petroleum
Building, built in 1958, and the Occidental Chemical Center, completed in1980. Much
nlore recent U.S. examples of double skin facades include the Manulife Building in
Boston (Figure 3.7) and the Genzyme Building in Cambridge (Figure 3.8), both
completed in the very early 21 st century. Considering the increased interest in energyefficient architecture, it is expected that major tall buildings adopting double skin facades
will emerge in the U.S. soon.
3.4. Conclusion
Modern architecture means discontinuation of millennia-long traditional architecture.
Based on new materialltechnology, new architecture was created. Classical orders and
heavy masonry walls with punched openings were replaced by steel columns/beams and
transparent glass curtainwalls. Traditionally hand-crafted architectural elements were
replaced by mass-produced ones in factories.
The machine metaphor originally
introduced by Viollet-le-Duc and reemphasized by Le Corbusier was not just a metaphor
any longer.
Architecture became a machine itself, and accordingly new esthetic
58
expression was sought and generalized as a new style. This phenon1enon prevailed in the
field of architectural practice including tall building developments for decades.
Compared with revolutionary original technology, which introduced a new breed of
architecture, remedial technologies follow a typical evolutionary process. In terms of tall
buildings, this phenomenon can be well observed by comparing a 1930's study on the
most economic height of a tall building in Manhattan with the number of stories of the
200 tallest skyscrapers built up to the present. While engineering height means the height
structurally feasible on a given site, the most economic height is the one that would
produce the maximum percentage of net return. As Carol Willis noted in her Form
Follows Finance, at some point in the construction of every skyscraper, the law of
diminishing returns sets in, and rents for the additional stories do not cover costs. As
buildings go higher, additional cost for most building systelTIS,such as structural systems,
becomes greater than proportional. On the contrary, as buildings grow upward, the
rentable space is expanded less than proportional due to the encroachment on rentable
area by the structural and vertical transportation systems. This sets the most economic
height at a certain level.
According to Clark and Kingston's study in 1930 among
several different height schemes, 63 stories was the most economical choice for a
:~~. ~.~
... ~+t
.j.
,,,, ..
. ++;
..
,"
BU!LDING
Manhattan
HEIGHT
IN STORIES
in 1930
59
20
U
-c:
18
~~
16
~<1>
co ..c 14
o
c
en .12
>. .......
L...
CJ)
__
.......
en
<1>
CO
OIL...
10
8.
<1> 0
..eN
0>
~ C
4
2.
Z a
17
25
33
41
49
57
65
73
81
89
97 105 113
Nurrber of Stories
in the World
Manhattan skyscraper (Figure 3.9). Today, among the 200 tallest buildings in the world,
76% of them are within the range of 50 to 70 stories, showing that this range is still the
most economic height (Figure 3.10).34 There has not been significant change in the
economic height, implying that over more than seventy years technological evolution has
been very gradual, which is very typical for any remedial technology following an
epochal original one.
This phenomenon may look ironical because today's technologies are much more
complicated than the crude original one. (If the sophistication of technology were able to
be measured quantitatively, today's technology might be more than ten times
sophisticated than the original one.) However, the major difference between them lies
not in their sophistication but in their nature. Original technology is idea-driven and
remedial ones today are mostly science-driven, and that matters.
There have been many reactions to modem architecture, as was discussed earlier. From
the viewpoint of the interaction between technology and architecture, some architectural
trends, such as post-modernism and high tech architecture, are reactions to the gradual
phase of technological evolution after the breakthrough original technology.
Land value is also an important factor in deciding a tall building's economic height. However, even
considering the importance of land value (e.g. Manhattan vs. the average big city of the world), there is no
better explanation for the lack of significant change of economic building height than the gradual evolution
process of technology after the breakthrough.
34
60
Today's architecture, including tall buildings, can be understood only through recognition
of the dominance of cultural pluralism. However, it is also useful to recognize that
today's pluralism in architecture can be substantially understood as the result of an
intellectual and cultural "branching out" from the notion of a modem architecture based
on modern technology. In fact, viewed in this way, modem architecture may now be
entering its culminating phase through necessary accretion of remedial technologies.
Reactions to this phase may be understood as premature pursuance of new architecture
even when the typical follow-up evolutionary phase of technology after the initial
revolutionary one has not yet been finished. This viewpoint well explains many recent
"-isms" of short longevity. Another revolutionary technology might put an end to today's
pluralism and lead to a new breed of architecture. And that new breed of architecture
will be again followed by refining remedial technologies for its culmination, just as is
happening with today's architecture.
61
62
CHAPTER 4:
TECHNOLOGICAL LIMITATIONS OF TALL BUILDINGS ANDFUTURE DIRECTION
As has been discussed so far, there has been notable evolution of technology since the
emergence of tall buildings in the late 19th century. However, many technological
limitations still exist. Indeed this is inevitable because of not only the experimental
imperfect nature of original technology but also the conflicting perceptual dual
characteristics of any modern technology. The promising reliability of modern
technology enabled architecture to finally move away from traditional architecture
toward a more modern approach. However, the other side of this promise has been the
ephemeral nature of technology. This is because the expected rate of technological
evolution based on the promising attributes exceeds the actual rate of the evolution to
meet this expectation. Also, the incremental rate of technology-related human
requirements of modern society exceeds the actual rate of the evolution to meet these
requirements, making the status of technology ever imperfect.
The ephemeral nature and the promise of technology are in conflict in modern
architecture. In tall buildings, which generally require the most advanced and varied
contemporary technologies due to their very tallness and scale, this conflict is dominantly
observed as their technological limitations. The most critical technological limitations of
tall buildings can be found in their structures, facades, and vertical transportations. The
following sections primarily discuss the first two systems in more detail. The discussion
of vertical transportation systems is limited in this thesis, not because it is less important
but because it is less relevant to the theme of this research.
of wind loads, stiffer structures produce less lateral displacement. With regard to the
dynamic portion of wind loads, the response of not only windward direction but also
across-wind
63
LOCK -IN
V (WIND
VELOCITY)
Since the natural direction of structural evolution towards lightness will not be reversed,
more stiffness and more damping should be achieved with a minimum amount of
material. Recently, diagrid structures have been emerging as a new structural trend,
providing stiffness very efficiently in tall buildings. Considering its structural efficiency
and architectural significance as a newly emerging esthetics of tall buildings at the time
of this research, Chapter 6 of this thesis is devoted to the investigation of the structural
performances and architectural potentials of diagrid structures. Chapter 7 and Chapter 8
investigate the strategy of increasing damping through design integration between the
structural system and the facade system.
appropriate structural design strategies to minimize the technological limitations structural-motion-related problems - but also as potential architectural esthetic drivers in
tall buildings.
From the viewpoint of structural material's property, as Connor (2003) noted "the lag in
material stiffness versus material strength has led to a problem with satisfying the
serviceability requirements on the various motion parameters and indeed for very high
strength materials it is possible for the serviceability requirements to be dominant." For
instance, today structural steels are available from 24 to 100 ksi (170 to 690 MPa).
However, its modulus of elasticity remains nearly the same without regard to the change
of its strength (Figure 4.2). The change of production process or heat treatment gives
impact on the strength but not on the modulus of elasticity of steel. There has been some
increase in the modulus of elasticity of concrete (Figure 4.3). However, this increase is
relatively small compared with the increase of strength. Also, steel, as the tall-building-
originating material from the 19th century, is the more dominant structural material than
concrete in tall buildings.
12
80
800
11
100
10
.><
9 .
.......
80
i~"
0.
400
~
g
40
50
60
~600
8,
~.,
~>
70
700
~
]>
..
~
0..
40
'300
30
200
20
1100
10
(.)
3
i
Str3;n.harde,,'og
to
m))t..
Tt:"\silf
range
t1rfongth
1
1
0.005
0.010
Strain
0015
c, in./in.
0.010
0.015
0.001
0.002
0.003
Strain, in.lin. lmm/mml
VS.
0.004
development of fluorescent lamps and mechanical HVAC systems during the late 1930s,
the substantial increase of office depth - typically measured from the inner face of
exterior facade to the outer face of interior core walls - has become possible because the
dependence on exterior windows for natural lighting and ventilation has been reduced.
Thus, column-free open office spaces have become much more desirable than the earlier
fixed-partitioned ones.
Even though today's information flow through computer
networks makes column-free open plans less necessary, they are still preferable.35
This
preference for column-free open space requires long structural spans, requiring deep
beams or trusses, which increase story heights and consequently building heights. In
addition, higher demand for HV AC and information system requires deep ceiling space
and/or sometimes raised floors. For instance, while the story height of the Empire State
Building, whose typical span is about 20 feet, is 11 feet 6 inches, that of the Sears Tower,
whose typical span is 75 feet, is 12 feet 10.5 inches. This phenomenon related to spatial
requirement makes today's skyscraper structures much taller and thus more flexible
compared with the old ones if the same total number of stories are built.
4.2 Technological
Limitations
The exterior wall system is normally supported by the primary structural system of tall
buildings. As Korista (1989) noted that "the behavior at the interface between the
relatively weak, generally non-ductile, individually determinant exterior wall system
elements with the higher strength, more ductile, highly redundant structural frame
elements is of critical importance ... ," the differential deformation compatibility between
For a more in-depth discussion on the evolution of space planning in tall office buildings, see Abalos &
Herreros, "Tower and Office - From Modernist Theory to Contemporary Practice," Chapter 5, pp. 177 215.
35
65
the primary structure and facade system is one of the very challenging technical issues of
tall building developments, especially in terms of long-term serviceability.3 6 Since the
emergence of skeletal structure and the curtainwall concept, leading the current status of
modern architecture from the viewpoint of technology - thus, even though this new
design and construction strategy must be appreciated overall - numerous facade failures
have occurred, especially in tall buildings, due to the incompatible deformation
performances between their primary structures and facades. These are caused by the
relatively large various loads tall buildings should carry due to their tallness.3 7
Another critical consideration with regard to the exterior walls of tall buildings is their
fundamental function/performance. The exterior walls act not only as fagades, which
govern the esthetics of buildings to a great degree, but also as an environmental mediator
between the exterior and interior. Often, there have been conflicts between these very
subjective and relatively objective requirements. In many cases, especially before early
1970s OPEC oil embargo, the performance requirements of facades as an environmental
mediator have been sacrificed, resulting in more energy consumption. However, today's
profound shifts in design approaches toward sustainable architecture take energy-saving
issues more seriously than ever, leading to the technological challenge to produce more
effective and efficient facade systems. Considering this, Chapters 7 and 8 of this thesis
are devoted to investigating the potential of double skin facades, not only as an effective
environmental mediator but also as a possible structural damping mechanism. Through
design integration between primary structures and double skin facades, also pursued are
architectural esthetic potentials.
36 Michael
Kim specifies the four fundamental criteria of successful building design; function/performance,
esthetics, constructability, and long-term serviceability.
37For more detailed discussion on this, see Korista, "Exterior Faqade System - Building Structure System,
66
because they must occupy too much space at their lower levels to provide sufficient
lateral stability. Thus, in tall buildings, adopting the modern concept of design and
construction - frame structures and curtainwalls - is not an option but a necessity.39
The separation of roles regarding structural and facade systems, however, did not
accompany physical separation. In fact, these two distinct systems have always been
together with physical contacts for their mutual benefit. Because of this symbiotic
relationship between them, it is useful to observe the co-related development of these two
systems through the various aspects of the theory of coevolution. However, one must
always be careful about applying the theories of one discipline into another due to their
similarity. Coevolution is not a human affair, but rather describes what happens in
nature with regard to living creatures' function/performance. The judgment of the
esthetics of the possible physical changes resulting from coevolution is not within the
realm of the human. In contrast, the technological co-development of structures and
facades of tall buildings is a process done not by nature but by humans architects/engineers. In addition, since this co-development is not within the domain of
nature but within that of artifacts that encompass art, not only its functional aspect but
also its esthetic expression are of critical importance. Thus, the theory of coevolution
about the living creatures in the nature can be adopted for a clearer understanding,
of the primary
juxtaposed and connected together at the perimeter of a building. Two systems oncefully-integrated with serious limitations have been relatively separated with minimum
connections necessary to contribute to each other to overcome the limitations.
Facades
have been supported by the components of the primary structures at building perimeters,
and in many cases primary structures have been protected from ever-changing harsh
outdoor environments by facades. Based on these fundamental mutual contributions,
these two systems have (or at least should have) coevolved in order to more effectively
minimize the contemporary technological limitations and also towards each other's - in
turn a whole building's - better performance. As nature determines the direction of
coevolution, architects/engineers control the direction of co-development of the two
independent but very closely related systems. In both nature and architecture, one
common goal is in the function/performance
domain.
However, in architecture - a
discipline which is actively engaged in the interests of both technology and art - both the
functional effectiveness/efficiency and the successful esthetic expression of integrallydesigned systems have equal importance.
39 This explains well the traditional theory of the technological invention based on necessity and utility in
terms of tall building construction. In terms of tall building development, as George Basalla (1988) noted
"necessity is a relative term." "A necessity for one people, generation, or social class may have no
utilitarian value or may be a superficial luxury for another people, generation, or social class. A skyscraper
is not a structure to protect people from the vagaries of the weather. Necessity is not something imposed
by nature upon humanity but is a conceptual category created by cultural choice."
67
In tall buildings, the meaning of the design integration between their structural system
and facades system is of special importance. The location of building perimeters has
more structural significance in tall buildings than in any other building type due to their
very tallness, which means greater vulnerability to lateral forces, especially wind loads.
Also, the role of facade systems in terms of function and esthetics is more significant in
tall buildings. In most cases, tall office buildings are stacks of many very similar floors,
which have relatively simple space organization - generally open spaces between the
inner face of the exterior facade and the outer face of the interior cores with movable
partition walls.
fundamental building esthetic definer is conspicuous in tall buildings. Due to this special
importance of the structural system and facade system at the perimeter of tall buildings
and their inevitable physical connections, design integration, which corresponds to the
coevolution process in nature, is the approach to be pursued to minimize technological
limitations and, in turn, to lead to refined better performance and esthetics.
Due to the special importance in many ways of perimeters in tall buildings, it is expected
that many building components are congested at this location. For instance, from the
viewpoint of structural behavior, it is quite desirable to concentrate as much lateral loadresisting systems as possible on the perimeter of tall buildings where building facades
always are.
It depends on the
68
PART II:
TECHNOLOGY IN TALL BUILDINGS AND THEIR ARCHITECTURAL
IMPLICATIONS
69
70
neighborhood of about a second, whereas that of gust and vortex-shedding can be very
similar to the fundamental period of tall buildings, resulting in resonance condition.
Various design strategies are employed to reduce the effect of wind loads applied to tall
buildings. They can be divided into two categories, passive systems and active systems.
Passive systems have fixed properties, and, in order to be performed as intended, they do
not require energy, while active systems do need energy source to modify the system
properties against ever-changing loads. Thus, active systems are in general much more
effective than passive systems. However, due to their economy, passive systems are
more commonly used than active systems.
5.1.1. Tall Building Designs with Passive Systems
The first strategy of the passive system category is to increase the lateral stiffness of the
structure, which is also the most traditional approach. In order to obtain the necessary
lateral stiffness, introduced first were braced frames followed by tubular structures, coresupported outrigger structures, and more recently diagrid structures. In the early braced
frames, the braces - the main lateral stiffness provider - are generally constrained within
the interior cores, and serve only for structural performance.
Consequently, no esthetic
building perimeter columns through the outriggers that connect them. This basic
configuration often requires perimeter super columns and/or belt trusses at the outrigger
levels, and these elements of the outrigger system are sometimes incorporated into
building esthetics.
Tubular and diagrid structures locate their major lateral load-resisting components at the
building perimeters where building facades are. This performance-induced juxtaposition
71
naturally leads to integrative design approach between the structural system and fa<;ade
systenl. As a consequence, in tall buildings that employ these
types of structural systems, technological components and
architectural esthetic components of building facades are
inseparable - eventually the same. These circunlstances require
very intimate cooperation between architects and engineers.
It is not possible to be a master builder today in this culturally
As their primary structural system, the Shanghai World Financial Center employs a tubular structure, and
the 7 South Dearborn employs a core-supported outrigger structure below the first notch from the ground.
However, in these buildings, these primary structural systems were not explicitly utilized as esthetic drivers
of the buildings.
40
72
Figure 5.3:
Turning Torso
Figure 5.4:
Freedom Tower
73
frames are considered, the lateral stiffness of the twisted forms is not great as straight
forms.
I
<
d\~\C
...
/
C-"
Ix=
b4
12
:-
b4
12
12
12
I-
b'
12
Ix =
12
The third strategy of the passive system category is to use auxiliary damping devices.
This strategy can be further divided into two categories: 1) Installing energy-dissipating-
74
Tuned sloshing dampers use waving water n1ass as a counteracting inertia force generator.
Thus, this system is usually designed using the existing water source in tall buildings,
such as a pool or water tank located at the top of a building. Sloshing frequencies are
tuned by adjusting the dimensions of the water container and the depth of water.
Another strategy of the passive systen1 category is to increase building n1ass to reduce
acceleration, which is, in general, the measure of n10tion-induced hUll1an discon1fort.
However, this strategy is not feasible due to its practical limitation of increasing building
mass only for the acceleration control. In addition, in case of earthquake, added mass
increases seisn1ic loads.
'i-=--f'~
j
1.....
Pro ......
aalanoK __
-"1
n.~.(IZl
75
Application
Method
TypeMeans Method
Type
Means
Passive Primary
Structure
Design
Application
Braced Frames,
Tubular Structures,
Core-Supported
Outrigger
Increase stiffness to
reduce lateral
displacement and
probability of
vortex-sheddinginduced lock-in
condition
Increase building
mass to reduce
acceleration
Aerodynamic
Design
Considerations~~~~~~~~~~~~~~~~~~~~~
Considerations-
Structures,
Diagrid Structures
Increase Building
Material
Chamfered
Comers,
Streamlined Form,
Tapering,
Improve
aerodynamic
properties to reduce
wind force
Not practical to
increase mass,
Amplifies
seismic loads
Reduces
occupiable
building areas
Openings, Notches
Auxiliary
Damping
Device
Viscous Damper,
Visco-Elastic
Damper, Hysteretic
Damper
Tuned Mass
Damper, Tuned
Liquid Damper,
Impact Damper
top of a building
Active Mass
Damper
More expensive
Generate
Aerodynamic
More expensive
aerodynamic
Appendage
than passive
control system
Active Variable
Stiffness System
More expensive
than passive
control system
counteracting
inertia force
Generate control
Active
Occupies
valuable building
than passive
control system
response
control force to
reduce wind force
Change stiffness to
avoid resonance
condition
.
76
dampers, are effective only for a narrow range of loading conditions, active systems can
perform effectively over a much wider range and they are a more advanced form of
functional performance-driven technologies in architecture, and in most cases they
remain in the performance domain. Examples are active mass dampers, active variable
stiffness systems, etc. The systems discussed so far are summarized in Table 5.1.
77
78
A simple methodology
member sizes of diagrid structures is presented. The optimal angle study and the simple
member sizing methodology will be useful at the early stage of diagrid structural design
for both engineers and architects interested in utilizing this structural system. Also
discussed are the architectural, constructability, and urban contextual issues of diagrid
structures.
Based on this research, it is expected that structural and architectural
decisions at the early stage of design can be made in a more integrative way and
efficiently.
6.1. History and General Characteristics of Diagrid Structure
The early designs of tall buildings in the late 1 9 t century
recognized the effectiveness of diagonal bracing members in
resisting lateral forces. Most of the structural systems deployed for
early tall buildings were steel frames with diagonal bracings of
various configurations
1\ I/
However,
/ I
1\ I/
/ \
Braced Tube
- A7--7
Structure
79
80
Mx~
- ,
Y1~
I
--
[a' i"
......
Ij
B
1--------
II
- - 4
members carry only axial forces. The cross section shear force is related to the diagonal
member forces by
V = 2Fd cos 0.
(6.1)
Assuming linear elastic behavior, the member forces are also related to the diagonal
extensional strain, Ed , by
d
(6.2)
AdEd
d~,
The extensional strain due to the relative lateral motion between adjacent nodes is a
function of A,, and 0:
Ah cos
ed
Ld
Ah cos 0 sin
(6.3)
sin 0
Neglecting the extensional strain in the diagonal due to rotation, A,
and approximating
Z A
(6 .4)
h
we obtain the following approximation for the total extensional strain:
E,
ycos
sin0
ysin20
2
(6.5)
Combining the above equations results in the following expression for this shear force,
82
__
V = (AiE
_ _
sin 2cos
O)y.
(6.6)
By definition, the transverse shear rigidity relates the shear force and shear strain,
V = Dy.
(6.7)
It follows that
D.7 = AxE sin 20cos 0.
(6.8)
In order to verify these assumptions and also to find the actual range of optimal angles, a
set of 60-story buildings having various diagrid angles are designed and analyzed using
SAP2000. The study is repeated for 42- and 20-story buildings. Details are presented in
the following sections.
0.6-
0.4 .
0.2-
. 20.
.
20
40
40
0
60
... 0
80
uIn
cosu
theta
generate different fundamental periods in the x and y directions, while diagrids are
configured symmetrically in both directions.
83
Description
Height
Width
Ext. Wall to Core Wall Distance
Core Area
Story Height
Beam Span
Beam Spacing
Floor Slab
Floor Live Load
Value
240m
36 m (square plan)
9 m typical
25 % of floor area
4 m typical
9 m typical
3 m typical
12 cm Concrete (wi WWF) on Metal Decking
50 psf
For the diagrid members, custom-made grade 50 steel pipes of varying diameters
inches to 15 inches are used. Their thickness varies from 3 inches for the
diameter pipes to 1.5 inches for the IS-inch diameter pipes. Box shape grade
built-up sections of varying dimensions are used for the core columns. Grade
sections are selected for the beams, and W 18 sections for the core bracings.
from 33
33-inch
50 steel
36 WI6
The document SEI/ASCE 7-02 (Minimum Design Loads for Buildings and Other
Structures) is used to establish the wind load. The buildings are assumed to be in Boston
and within category III, which implies that there is a substantial hazard to human life in
the event of failure. Based on the code, the basic wind speed is 110 mph. Lastly, one
percent damping is assumed for the calculation of the gust effect factor.
84
Two different diagrid structural schemes are examined. Scheme 1 includes four vertical
comer columns; scheme 2 has no vertical comer columns. The same set of 7 different
diagrid angles (Figure 6.10) is considered for both schemes, resulting in 14 different
building structures. Scheme 2 requires slightly larger members and uses about 10% more
steel than scheme 1. For instance, for scheme 2, the diagrid members at the ground level
are 33-inch diameter pipes, while only 27-inch diameter pipes are required for scheme 1.
Member sizes were generated for the building with a diagrid angle of 63 degrees by
iterating on the sizes until the deflection constraint for the top floor,
u(H) ~ = 0.48m
500
was satisfied. This design process is based on the assumption that motion constraints
rather than strength constraints control the design of tall buildings. The design was
checked for strength using the LRFD code and found to be acceptable. A value of 63
degrees was taken since our preliminary study indicated that it was close to optimal.
These member sizes were also used as initial estimates for the other 6 diagrid systems
corresponding to different diagrid angles. Calculations indicated that the lateral stiffness
is not too sensitive to angle in the region of 63 degrees. Thus, the member sizes were
slightly increased so that the structures with angles of 53, 69 and 76 degrees also satisfied
the maximum displacement requirement.
Strength checks were also made on these
structures. Figure 6.11 contains a plot of the lateral displacements for the scheme 1
structures. The results indicate that angles between 53 and degrees and 76 degrees are
reasonable choices.
85
Horiz. Dis I.
0.63
0.47
0.42
0.43
4.53
3.84
3.66
3.71
4.10
0.50
5.45
7.95
0.87
1.88
2.00
1.80 1.60
1.40 -
1.20 -
L.
Q)
Q)
1.00
@ Top (m)
0.80 0.60
0.40 0.20 0.00
34
53
63
69
76
82
90
degree
A similar process was used to select member sizes for the scheme 2 structures. Member
sizes were generated to satisfy the maximum displacement requirement for the structures
with angles of 63, 69 and 76 degrees in this case. Displacement results for the scheme 2
structures are plotted in Figure 6.12. Since all the vertical columns that provided some of
the bending rigidity have been removed, the diagrids need now to provide not only all the
shear rigidity but also all the bending rigidity. Thus, the angle range is shifted upward,
and centered on about 70 degrees.
Results are summarized in Table 6.2 and 6.3. Comparing Table 6.2 and 6.3, one observes
that for diagrid structures with small angles, the vertical displacement is higher than for
conventional structures with vertical columns.
However, this problem is not so
significant.
86
Horiz. Dis I.
7.44
4.55
1.67
0.66
3.86
3.73
3.96
0.47
0.44
0.48
5.18
8.21
0.79
2.00
I~
34
53
63
69
76
82
90
degree
87
lateral stiffness is provided only by the diagrids for the 42 and 20 story structures
presented later in this chapter.
Figure 6.13: Anatomical Models of Oiagrid Structure for Relative Stiffness Assessment
6.00
Relative Stiffness K
5.00
~
U)
4.00
U)
Q)
c:
~U) 300
Q)
>
:;:;
Diagrids ..K
__
Braced CoreK
.!!! 2.00
1.00
0.00
I
L~
34
53
63
69
76
88
82
90
degree
Figure 6.15: Scheme 1 (With Corner Columns) and Scheme 2 (Without Corner Columns) Models
89
I~
CI)
E 0.60
0.40-
0.20
0.00,
34
53
69
63
76
82
degree
l __
--.
0.70
_._-_ _~_
..
~"
.......
0.60 0.50 -
Q; 0.40
a;
E 0.30
I ~Horiz.
0.20 0.10
0.00 .
34
53
63
degree
90
69
76
82
Displ.@Top(m>!
In this section, a simple methodology for the preliminary design of diagrid structures is
introduced. As was discussed earlier, typical braced cores contribute only 15 to 20% of
the total lateral rigidity in properly designed diagrid tall buildings. For this reason and
because we can more accurately estimate the lateral rigidity provided by diagrids, all the
required lateral stiffness is allocated to the diagrids.
The procedure begins by specifying the contributions to the total lateral displacement of
the structure due to bending and shear deformation. This allocation is based on an
assessment of the intrinsic attributes of diagrid structures and the behavioral
characteristics of tall buildings related to their height to width ratio, and leads to the
required values of shear and bending deformation. Simple equations, which convert
these required deformation values into cross sectional areas for the diagrid members, are
derived.
This design methodology is applied to several diagrid structures of various heights and
aspect ratios. Based on these studies, empirical guidelines for assessing the relative
contribution of bending and shear deformations to the total lateral displacement of diagrid
tall structures are derived. With this formula, the process is essentially automated.
LOADING
DIRECTION
91
diagonal members are assumed to be pins-ended, and therefore to resist the transverse
shear and moment through only axial action. With this idealization, the design problem
reduces to determining the cross-sectional area of typical web and flange members for
each module. These quantities are established with a stiffness based approach (Connor,
2003).
The shear force, V, and bending moment, M, are expressed in terms of the relative
displacement and rotation measures, Au and Af, for the module as
(6.9)
V = KAu
(6.10)
M = KA,8i.
The motion measures
measures by
Au =y h
(6.11)
A, = .
(6.12)
K-
2N
(W21V
,EL,
K =N 2
Cos2KJ
(6.13)
COS)
(6.13)
d,+"
Lasin2
(6.14)
where Nv is the number of diagonals extending over the full height in one web plane, and
Nf is a similar count for one flange plane.
Given V and M, one specifies the desired transverse shear and bending deformations, y
and X, and determines the required stiffness using Equations 6.9 and 6.10:
K 7.=
92
y h
(6.15)
(6.16)
K, = -M
Lastly, substituting for the stiffness terms, one obtains expressions for the typical areas in
the web and flange:
VL
=,
VL
2N,,E hy cos2 0
A, t
2MLj
NB2Ehsin
2
NB E,,Xhsin
22 00
(6.17)
(6.18)
Since the diagonal members are assumed to be constant in a plane, one needs to consider
loading in both directions in order to establish an upper board value for the areas.
*H2
u(H)=y H+ 22(6.19)
where y H is the contribution from shear deformation and X*H2 /2 is the contribution
from bending. In order to specify the relative contribution of shear versus bending
deformation, we introduce a dimensionless factor s, which is equal to the ratio of the
displacement at the top of the structure due to bending and the displacement due to shear:
S=(/
_(y*H)
'X*
(6.20)
u(H) =--
(6.21)
93
u(H) = (1 +s)h =
(6.22)
Then,
7*
(6.23)
(1+s)a
2s
H( + s)
2 *s
H
(6.24)
Typical values for a are in the neighborhood of 500. It remains to establish a value for s.
We introduce a dimensionless factor, f, which is defined as the ratio of the strain in a web
diagonal due to shearing action to the strain in a flange diagonal due to bending action.
(6.25)
ld,'eb
ed,flange
d,'we
AucosO
Ld
cd,flunge=
BAri sin 0
21s
Au cos sinO
h
= cossin
BAri sin 20
2h
Ld
=s
2 Xsin2 0,
(6.26)
(6.27)
fl L/,ge
2ycos sin
BX sin2 0
2y
B% tan 0
(6.28)
Combining Equation 6.28 and Equation 6.24 leads to an equation for s in terms of f, the
HX
2y
94
H(2 7 Bf tanO) _ H
Bf tan0
2y
f
I
allowable incremental
deformation in the
inelastic behavior.
However,
H/B
S
in diagrid
H/B
Figure 6.19: fvs. s
The methodology is applied to 60-story and 42-story diagrid structures with an angle of
63 degrees.
The first step is to divide the structure into appropriate structural modules.
For the 60-story building, every 6-story segment is used as a structural module in this
example. For each structural module, shear forces and bending moments are calculated
using the code loadings described in Section 6.2.3. The 60-story building has an aspect
ratio of about 7, and 'f was taken equal to 0.5. The corresponding value for s is about 6.
Deformation measures were based on a maximum lateral displacement of H/500 = 0.48
meters. Member sizes for the modules were computed using the following equations
customized for the 6-story module shown in Figure 6.18.
Ad
Ldcos 2
,,Ejhycos
1\2N
12E VLcos
12Ehycos
(6.30)
95
dJ
2MLd
N.rB2Edxhsin2{)
2MLd
(6+2)B2Edxhsin2{)
(6.31)
An estimate of the contribution of the diagonals on each web to the bending rigidity is
made by adding one extra diagonal on each flange, resulting in N.r ~ 6 + 2. Profiles of
the required areas for the typical diagonals in the web and flange planes are plotted in
Figure 6.20. Since the wind can blow in either direction, the role of a plane can be either
a flange or a web. The building considered here has a square plan and the preliminary
design value for the module is taken as the larger of the two values (bold face in Table
6.4)
V(KN)
Story
55th - 60th
M(KN-M)
7,950
2,650
5.300
49th - 54th
43rd - 48th
37th - 42nd
31st - 36th
25th - 30th
19th - 24th
71,550
198,750
7,950
10,442
389,076
639,417
12,845
15,248
947,697
1,312,848
17,384
18,986
13th - 18th
7th - 12th
1st - 6th
Ad(bndg)
0.0008
0.0069
0.0191
0.0374
0.0615
0.0911
0.1262
Ad(Shear)
0.1662
0.2100
0.2681
0.1483
0.1608
0.1734
1,728,462
2,184,126
20,588
22,190
2,788,776
0.0207
0.0414
0.0621
0.0816
0.1004
0.1191
0.1358
Table 6.4: Preliminary Member Sizing for the 60-Story Diagrid Structure
0.3000
0.2500
0.2000
M for Shear
-II-
oS
Q)
E 0.1500 C"
t/)
0.1000
0.0500
0.0000
~~
<0
~
~~'
Oj~'
'b~
~
t!:>\~'
A~'
~\
~~
~
q..'
.....
fo~
fo~'
t);:
~~
'b~
'V
Oj~'
p.,,:
"
~
~~'
'\~'
"
"q..
stories
Figure 6.20: Preliminary Member Sizing for the 60-Story Diagrid Structure
96
A similar study is carried out for a 42-story structure with an aspect ratio of 5 and angle
of 63 degrees. The parameters for this case are f = 1 and s = 2. Diagonal areas predicted
by Equations 6.30 and 6.31 are plotted in Figure 6.21, and the suggested design values
are listed in Table 6.5 (bold face).
Story
37th - 42nd
31st - 36th
25th - 30th
19th - 24th
13th - 18th
7th - 12th
1st - 6th
V(KN)
2,492
4,895
7,298
9,434
11,036
12,638
14,240
M(KN-M)
7,476
67,017
184,497
358,848
Ad(bndg)
0.0006
0.0058
0.0160
0.0311
Ad(Shear)
0.0505
0.0734
0.1092
0.0370
0.0423
0.0477
583,662
848,526
1,262,376
0.0083
0.0164
0.0244
0.0316
Table 6.5: Preliminary Member Sizing for the 42-Story Diagrid Structure
~hi
for Bending
0.1200
0.1000
0.0800 -
SQ)
E 0.0600
ern 0.0400
0.0200 0.0000 -
Structure
97
The structures having these "preliminary" design areas were analyzed with SAP2000 to
assess the accuracy of the preliminary sizing methodology. Table 6.6 compares the
results for the displacement at the top of the structure and the required steel tonnage with
the conventional approach. There is reasonably close agreement. Considering that no
iteration is required, we conclude that the methodology is useful for member sizing at the
Structure
Umax (Target)
Steel
Method
Umax
Proposed Method
Conventional
Method
Proposed Method
Conventional
Method
0.30m
0.29m
1,620t
1,850t
0.43m
0.47m
6,080t
5,440t
Tonnage
42 story
168 m
36 m
1.0
0.34 m
60 story
240 m
36 m
0.5
0.48 m
set of diagrid structures with heights ranging from 33 to 69 stories and various aspect
ratios were considered.
Figure 6.22 shows sets of diagonal areas for the 60-story building, having an aspect ratio
of (H/B) of 6.7, and diagonal angle of 63 degrees. When s is taken as 1, the maximum
displacements at the top due to bending and shear are equal, and the areas required to
limit displacement due to bending govern the design for square plan buildings (Figure
6.22 a). When s = 9, the maximum displacement at the top due to bending is 9 times that
due to shear, and the areas required to limit displacement due to shear govern the design
for square plan buildings (Figure 6.22 b). These cases represent the extreme limits for s.
Choosing s to be 4 and selecting the greater of the member sizes required for the bending
and shear criteria, leads to the most economical design (Figure 6.22 c). In this case, the
bending deformation requirement governs for the lower half of the building, and shear
deformation requirement for the upper half.
Table 6.7 contains a comparison of the targeted maximum displacement and the
displacements computed with SAP2000. The results show that all the designs are
conservative, and the designs corresponding to s values of 3, 4 and 5 appear to be closer
to optimal. A closer estimate can be obtained by comparing the steel tonnage required
for each design. These results are plotted in Figure 6.23, and indicate that values of s
between 3 and 5 are equally optimal.
98
l!! 0.3000
S
0.2500
a-
0.2000
III
':)
rS:>~
~_b<~
'V
'J
~'~'b<
'b~
~
#'~'~'
~\
b<
~'6
b<~
~
ro"S'
~
~
t;:)~
,~
"
~~
,t');
"
'b~
,"
~
,"
~
~:
"
stories
(a)
l!!
S
E 0.1500
Q)
a-lii
0.1000
0.0500
(b)
0.3500
0.3000 0.2500 -
.s~
0.2000
Q)
a-
0.1500
III
0.1000
0.0500
0.0000 55th - 49th - 43rd - 37th - 31st - 25th - 19th - 13th - 7th - 1st60th 54th 48th 42nd 36th 30th 24th 18th 12th 6th
stories
(C)
Figure 6.22: Preliminary Member Sizing for the 60-Story Diagrid Structure with Various s
99
1
2
3
4
5
6
7
8
9
Umax (tarQet)
Umax
m
m
m
m
m
m
m
m
0.48 m
0.48
0.48
0.48
0.48
0.48
0.48
0.48
0.48
(SAP2000)
0.35 m
0.42 m
0.45 m
0.45 m
0.45 m
0.43 m
0.41 m
0.39 m
0.36
Similar studies were carried out for 33-, 42-, 51- and 69-story diagrid structures with
aspect ratios ranging from 4 to 8. Table 6.8 and Figure 6.24 contain the results for the
set of 42-story structures. All the designs are conservative, and the optimal range for s
appears to be between 2 and 3.
Based on these studies, the following empirical relationship between the optimal s value
and the aspect ratio is proposed for diagrid structures greater than 40 stories with an
aspect ratio greater than about 5 and a diagrid angle between 60 and 70 degrees.
H
-~5
B
100
(6.32)
2
3
4
Table 6.8: Umax
Umax
(tarQet)
Umax
0.34 m
0.34 m
0.34 m
0.34 m
(SAP200Q)
0.25 m
0.30 m
0.28 m
0.26 m
101
for H/250, some of the diagrid members failed the code requirements.
failure occurred on the leeward face of the structure.
The same study was repeated for the 42-story diagrid structure with s = 2. Figure 6.26
shows the member sizes for 42-story diagrid structures based on two different maximum
displacement criteria, H/500 (0.34 m) and H/400 (0.42 m). It was found that, for H/400,
some of the diagrid members fail the code requirements. Again, the most severe failure
occurred on the leeward face of the structure.
Figure 6.25
0.1000
0.0800
l!?
Q)
Qi
E
C-
eil
0.0200
37th42nd
Figure 6.26
102
31st36th
25th30th
19th24th
13th18th
7th12th
1st-6th
103
6.7. Conclusion
The influence of the diagonal angle on the behavior of diagrid type structures was studied.
The schemes presented earlier in this chapter are several design alternatives for both
architects and engineers. Subjective contemporary esthetic choices from those design
alternatives having various diagrid configurations may or may not coincide with optimal
structural performance. Thus, design decisions should be made based on the recognition
of this fact.
For 60-story diagrid structures having aspect ratio of about 7, the optimal range of diagrid
angles is from about 65 degrees to 75 degrees. For 42-story buildings having aspect ratio
of about 5, the range is lower by around 10 degrees because the importance of bending to
the total lateral displacement is reduced as the building height decreases.
A stiffness-based methodology for determining preliminary sizes for the diagonals was
introduced and applied to a representative set of steel buildings. Results for displacement
and required steel tonnage demonstrate the practical usefulness of the proposed
preliminary design method. Compared with a conventional strength-based iterative
methodology, a stiffness-based methodology is more efficient for today's relatively light
and flexible structures such as tall buildings, the design of which is in many cases
governed by motion rather than strength.
The importance of interactions between architects and engineers based on mutual
understanding is emphasized through this thesis to produce better architecture. Based on
this recognition, emphasis was placed on developing relatively simple methodology that
is easy to follow. This methodology will be accessible not only for engineers but also for
architects so that they can communicate better toward more integrative design.
It is expected that the optimal angle study results, the simple member sizing methodology,
and other topics discussed, such as architectural, constructability and urban contextual
issues, will be very useful to both architects and engineers for preliminary design of
diagrid structure. Based on these studies, structural and architectural decisions at the
early stage of design can be made in a more integrative and efficient way.
104
CHAPTER
7:
STRUCTURAL
DYNAMIC
Low STIFFNESS
MOTION
DSF CONNECTORS
CONTROL
USING DOUBLE
AS DAMPING
SKIN FACADES:
MECHANISM
Double skin facades were introduced, earlier in this thesis, as a remedial technology that
remedies the problematic aspect of single skin facades as an environmental mediator.
Double skin facades create new space between the inner and outer skins. This depth not
only contributes to enhanced environmental control but also creates augmented
perceptual transparency as well. Since the emergence of this space is relatively new, its
functional and esthetic potential is not well defined yet. Thus, more possibilities of this
new additional space, introduced by new technology, are explored in this and the next
chapter.
7.1. Introduction
to the System
Tall buildings, which emerged in the late 19th century in the U.S, were a
so-called "American Building Type," meaning that the most important
modem tall buildings were built in the U.S. Today, however, they are a
worldwide architectural phenomenon.
Many tall buildings are built
especially in Asian countries, such as China, Korea, Japan, and Malaysia.
There has been some skepticism regarding constructing tall buildings
since September 11, 2001. Nonetheless, due to their significant
economic benefits in dense urban land use, tall buildings are and will be
built continuously all around the world.
Figure 7.1:
Various fa<;ade systems, such as glass/metal curtainwalls, precast
concrete panels, and stressed steel skins, are used to clad tall buildings.
DSF
Generally, most fa<;ade systems are composed of several layers. In
conventional cases, there are no substantial gaps between the fa<;ade layers. The double
skin fa<;ade (DSF) system (Figure 7.1), which has a substantial cavity between the fa<;ade
layers, has been obtaining increased interest especially in Europe mainly due to its energy
efficiency by enhanced performance as an environmental mediator. Through the cavity,
for example, hot air can be effectively removed in summertime, and also natural
ventilation can be introduced even at tall buildings' higher levels because there are
additional exterior skins that act as wind buffers. These functions of the DSF system
reduce energy usage in building operation, resulting in economic benefits in the long run,
even though their initial construction cost is higher than conventional single skin fa<;ades.
While many studies have been done regarding environmental/energy aspects of the DSF
system, no research has been done on the structural capability of the system. In tall
buildings - especially at their upper portion - excessive movement/acceleration can cause
serious human discomfort.
This chapter investigates tall building dynamic motion
control by introducing energy dissipating mechanisms within the DSF cavities.
Furthermore, their architectural implications are discussed.
105
Two schemes are studied. The first one introduces low stiffness
connectors in the DSF system (Figure 7.2). The concept is designing
the connectors between the DSF outer skin and the building's
primary structure to be relatively soft so that the transmissibility of
the dynamic portion of wind load can be reduced through them. As
a result, the DSF outer skin moves back and forth substantially, but
the vibration of the primary structure, which is enclosed by the inner
skin and contains occupants within it, is reduced significantly.
Dynamic motion control for tall buildings is achieved through this
mechanism.
However, one challenge in this scheme is how to
minimize the outer skin's vibration, which may involve serious
constructional,
visual and psychological
concerns, without
sacrificing effectiveness of the system.
The second scheme is investigated to overcome the challenges
confronted during the study of the first scheme. In this scheme, the
DSF outer skin is fixed as in conventional DSF systems. While the
DSF outer skin mass itself, attached to soft connectors, is used as a
counteracting inertia force generator in the first scheme, additional
small masses are inserted into the DSF cavity in this scheme in order
to act as distributed tuned mass dampers (TMDs), which effectively
control tall building vibration under dynamic loads such as wind
(Figure 7.3). Even though the problems of the first scheme are
eliminated, the disadvantage of this scheme is that this design
requires additional mass, which was not necessary in the first scheme.
However, compared with a conventional TMD system, which is
usually located near the top of the building, this scheme has the
significant benefit of saving valuable occupiable space.
Primary
Structure
f-1
~~
:
:
--
:
(
~-:
I
I
I
I
I
I
I
I
II
I
I
-I
I
I
I
I
I
I
-I
-I
I
I
I
I
I
106
Both the first and second schemes are investigated in more detail through the rest of this
thesis. For the low stiffness DSF connector scheme, the system is analyzed using a
simplified two degrees of freedom (2DOF) system model.
Its effectiveness and
limitations are discussed based on the results of the analyses of this system. For the
vertically distributed TMD scheme, system analysis theories are introduced, and a 60story building is simulated using a 60DOF model with vertically distributed TMD system.
The result is compared with a conventional TMD scheme, and the advantages and
disadvantages of the proposed system are discussed in greater detail.
7.2. Low Stiffness DSF Connectors Subjected to Periodic Loading
-;'
=,
~
=>
=>
:J
---'-
=;>
-""
-;'
---'-
M: Mass
K: Stiffness
C: Viscous Damping Parameter
P: Applied Loading
U: Displacement
0): Natural Frequency
of Structure
md: Natural Frequency of DSF Outer Skin
Q: Forcing Frequency
p: Forcing Frequency to Natural Frequency Ratio (rum)
f: DSF Outer Skin Frequency to Primary Structure Frequency Ratio (md / m)
~: Primary Structure Damping Ratio
~d: Fa9ade Connector Damping Ratio
The
107
p pe
pe=ip
-
(7.3)
(7.4)
l-=
Uein
(7.5)
where the response amplitudes, and u,, are considered to be complex quantities.
Then the corresponding solution is given by either the real or imaginary parts of u and u,.
Substituting Equations 7.3, 7.4 and 7.5 in the set of governing Equations 7.1 and 7.2
results in
- Q2mU + ifcu + k = iQcdud + kdUd
(7.6)
(7.7)
(7.8)
c = 2com
(7.9)
2 k
d
C =
dWImd
)
(7.10)
(7.11)
(7.12)
defining 'f as the DSF outer skin frequency to primary structure frequency ratio,
f =JWctI
0
(7.13)
and defining 'p' as the forcing frequency to primary structure frequency ratio,
P=Q/(0
the
governing Equations 7.6 and 7.7 bcome
the governing Equations 7.6 and 7.7 become
108
(7.14)
(7.15)
(7.16)
Then, the solutions due to a harmonic excitation can be expressed as
- =-ne
P
k
LT
io,l
(7.17)
(7.18)
Here, H is the dynamic amplification factor of the primary structure, and Hd is that of the
DSF outer skin. Figure 7.5 contains the plots of Hand Hd with various DSF connector
damping ratios. The DSF outer skin mass is assumed to be 1% of the primary structure
mass. Soft connectors between the primary structure and the DSF outer skin are
employed to reduce the transmissibility between them. Frequency ratio f (DSF outer skin
frequency / primary structure frequency) is 0.5, and the primary structure damping ratio
is 5% in this study. With this 5% structural damping ratio, the maximum H value of the
primary structure without the DSF energy dissipating mechanism is about 10.
H
Hd
2500
2000
1500
1000
500
0.2
0.4
0.6
0.8
rho
DYNAMIC AMPLIFICATION
FOR PRIMARY STRUCTURE
1.21.41.61.8
FACTOR
(H)
H olot . U= (P/K) H
Mass Ratio (Md/M) = 0.01
Structure Damping Ratio = 0.05
Fa<;ade Damping Ratio = 0.08, 0.2, 0.4, 0.6, 0.8, 1.0
f (Primary Struc. Period /Fa<;ade Period) = 0.5
0.2
0.4
0.6
0.8
1.2
1.4
1.6
1.8
rho
DYNAMIC AMPLIFICATION
FOR DSF OUTER SKIN (Hd)
FACTOR
Hd olot . U= (P/K) Hd
M.R. (Md/M) = 0.01
S.D.R = 0.05
F.D.R. = 0.08, 0.2, 0.4, 0.6, 0.8, 1.0
f= 0.5
Figure 7.5
As can be seen from Figure 7.5, when the fa~ade connector damping ratio is 0.2, the
maximum H value is reduced about 50%, compared with the case when there is no DSF
energy dissipating system. As the fa~ade connector damping ratio increases above 0.2, H
109
value increases,
of the primary
connectors. As
p = 0.5. When
eventually
structure)
the fa~ade
the fa~ade
the fa~ade connector damping ratio is 0, the primary structure resonates at p = 0.5.
With regard to the vibration of the DSF outer skin, higher fa~ade damping ratio is always
better to reduce Hd.
In general, the vortexshedding-induced
lock-in
phenomenon produces the
0.22 HZ
.~
most
severe
vibration
problem.
When 20%
0.17 HZ - - - - - - - - - - - - - - - - - - - - - - - - - ~
'
fa~ade connector damping
( 8LOO FREO..9CY)
~/ LOO< -IN :
ratio
is considered
to
minimize the motion of the
0.00 HZ
primary structure, at p =
- - - - - - - - - - - - ~-IN
( SKIN FREO..9CY )
0.5, the fa~ade lock-in
condition will occur, even
in moderate wind speed
V(WlIDVROOlY)
100M=tl
since the natural frequency
of the fa~ade connector is
half
of
the
building
Figure 7.6: Primary Structure and Fa~ade Lock-In Condition
frequency.
Under this
condition, as can be seen from Hd plot in Figure 7.5, dynamic amplification factor of the
fa~ade, Hd, becomes very large.
I
'----
---L-
..L.-_~_
For example, for a 40m x 40m square plan tall building, when wind velocity varies from
20mph to 100mph, the vortex shedding frequency fv varies from 0.04 to 0.22, following
the approximate Equation 7.19 and assuming the Strouhal Number as 0.2.
fv= VS/D
(7.19)
V: Wind Velocity
S: Strouhal Number (Reynolds Number & Building Geometry dependant)
D: Across-Wind Dimension of the Building Plan
Considering the approximate natural period of tall buildings (about 60 stories) to be
approximately 6 seconds, this frequency as well as the low stiffness DSF connector
frequency is within the possible vortex shedding frequency, resulting in lock-in
conditions. The primary structure lock-in condition will occur around 80 mph wind
speed, and fa~ade lock-in condition around 40 mph wind speed when f = 0.5 (Figure 7.6).
During the primary structure lock-in condition, the dynamic response of the primary
structure will be reduced due to the low stiffness DSF connectors. However, during the
fa~ade lock-in condition, the DSF outer skin will vibrate severely. In fact, in this case,
the DSF outer skin acts as a force amplifier for the primary structure.
110
7.3. Conclusion
III
112
CHAPTER
8:
STRUCTURAL
DYNAMIC
VERTICALLY
DISTRIBUTED
8.1. Introduction
MOTION
TMDs
CONTROL
USING DOUBLE
SKIN FACADES:
to the System
Tuned mass dampers are generally used to reduce wind-load-induced dynamic response
of tall buildings. TMDs are usually located near the top of tall buildings because they
work most effectively when they are located where maximum lateral displacement occurs.
However, as a result, very valuable top floor space is sacrificed merely as TMD room.
ELEVATION
SECTION
PLAN
Figure 8.1: Conceptual Sketch of Vertically Distributed TMDs within DSF Cavities
As mentioned in Chapter 7, there are several benefits of distributing TMDs through the
DSF cavity. First of all, by doing this, no occupiable building area is used as a room for
conventional huge TMDs. Second, since all TMD masses are located at the extreme
perimeter of the building, torsional resistance exerted by TMDs is maximized if they are
designed to move in the direction parallel to the building fayade planes. Third, by
distributing TMDs vertically not only the first mode, which has its maximum
displacement at the top, but also other modes, which may not have their maximum
displacement at the top, can be effectively controlled if necessary. In addition, TMDs
can be more easily installed by making them small. Figure 8.1 shows the conceptual
sketch of distributed multiple small TMDs within the DSF cavity.
113
8.2.1. Approximate
English/Greek
Svmbols
M: Mass
K: Stiffness
C: Viscous Damping Parameter
P: Applied Loading
U: Displacement
m: Natural Frequency of Structure
n: Forcing Frequency
p: Forcing Frequency to Natural Frequency Ratio (Wm)
f: TMD Frequency to Primary Structure Frequency Ratio (md
~u,
/ m)
P2
A 2DOF system with TMDs attached to both primary masses 1 and 2 is considered to
represent vertically distributed TMDs in a tall building. The governing equations for the
system shown in Figure 8.2 are
114
MU,
i +c,Il +kklu-
20(U,
(8.1)
- zU,)
- k2(U2 - UI) = I +kdIud + Cdlz~,1
m2 u2 + c (U2 -U, ) + k2 (u 2 - u ) = P2
+ kd
2ud2
(8.2)
+ Cd2d2
(8.3)
(8.4)
Equations 8.1 and 8.2 are combined to reduce the problem to an equivalent SDOF system
for the approximate solution. The approach followed here is based on transforming the
original matrix equation to scalar modal equations. Introducing matrix notation,
Equations 8.1 and 8.2 are written as
n,
Lii
m
CC
uii
H2
2
2
p'
MU+CU+KU=
P2
C2
U2
k +k 2
-
k 2/Lu
k2
2I
p2
P2
+{ kdludl + CdIUCldl
kd 2 ud,2 + Cd2H2
9
kd Udl + CdI
Cdl
{kd
2 ud 2
(8.5)
(8.6)
+ Cd2 Ud 2
{022
e J
'M,
k =
(8.8)
= mi
(8.9)
(8.10)
(8.11)
+c,
+klq = ,,11(p, +kdlUdl +Cdudl)+012(p2 +kd2Ud2 +Cd2Hd2)
n12q2 + C21
2 + k2 q2 =
21 (P1
+ kdlI
+ Cdlldl) +
22 (P2
(8.12)
(8.13)
115
The displacement at the top of the building is obtained by superposing the modal
contributions.
U2 =
12 ql
(8.14)
2 2 q2
However, in general, the first modal response is dominant in tall buildings subjected to
wind loads. When the external forcing frequency is close to the fundamental frequency of
the primary structure, co, the first mode response will dominate, and it is reasonable to
assume
U2
(8.15)
012ql
I2
IIPl +012P2
(12)2(012
2// )
(012)2
+ (0- )22
12
012
(8.17)
ole
2
(01 )
(8.18)
Cl
(8.19)
(012 )
(8.20)
P 1Pl
12P2
(8.21)
012
Taking u2 - , m1
116
l,
e
C
-, k
(8.22)
ii + c
(8.23)
(8.24)
=-aml,,uii
ifd2Ud9
+kkd2U
u 2 =-m,,u
md2u
2 +CJ9Ud2
+ C2ud, +
(8.25)
(8.25)
It is convenient to work with the solution expressed in terms of complex quantities. The
force is expressed as
p = pe'
where
(8.26)
= ue'
Ud =
(8.27)
'e
(8.28)
+ Cd2Ud2 + kd2Ud2 =
2U
-Q-m
2)
(8.29)
(8.30)
(8.31)
(8.32)
c = 2com
(8.33)
od = /n
(8.34)
c.j
= 2a dw,,mn,
(8.35)
(8.36)
and defining f as the TMD to primary structure frequency ratio,
(8.37)
the governing Equations 8.29, 8.30 and 8.31 become
(8.38)
(8.39)
(8.40)
The solutions due to periodic excitation can be expressed as
(8.41)
(8.42)
(8.43)
Figure 8.3 contains the plots of H , H dl' and H d2' when the mass ratio ml and m2 are
1.5%, and TMDs are tuned by iteration for optimal performance.
damping ratio was used for this plot.
9
B
7
6
5
4
3
rho
118
As can be seen from Figure 8.3, the H plot shows only two peaks, even though the
original 2DOF/2TMD system was converted to a 3DOF system. In fact, if TMD
damping ratio ~d = 0, we can observe one more peak between the two peaks. However,
this peak in-between is very sensitive to TMD damping ratio
~d.
very little damping for TMDs eliminates this peak. This phenomenon is generally
observed in vertically distributed TMD systems. In a NDOF+NTMD system, if the
NDOF primary structure is reduced to a single degree of freedom system, the H plot for
this system has N+ 1 peaks if ~d = 0 . However, once damping is applied to TMDs for
tuning, the H plot has only two peaks, and its overall shape becomes similar to that
shown in Figure 8.3.
8.2.2. Exact Solution for 2DOF/2TMD System
In the previous section the approximate solution was introduced in order to simplify the
analysis of the vertically distributed TMD scheme. In this section the exact solution is
studied: 1) to verify the accuracy of the approximate solution, and 2) to investigate the
contribution of the second mode.
Enl!lishlGreek Symbols
M: Mass
K: Stiffuess
c: Viscous Damping Parameter
P: Applied Loading
U: Displacement
m: Natural Frequency of Structure
0: Forcing Frequency
p: Forcing Frequency to Natural Frequency Ratio (n/m)
f: TMD Frequency to Primary Structure Frequency Ratio (md 1m)
1;: Damping Ratio
~: Mode Shape Vector
For this purpose, our beginning point is Equations 8.12 and 8.13, which are rewritten here,
from the previous section:
(8.12)
(8.13)
In the approximate solution, we neglected the second mode, and in turn, Equation 8.13.
In the exact solution, it is included. Through the very similar process presented in the
previous section, Equation 8.12 can be rewritten
(8.44)
119
1(i2dfimIPl + f2m-i),d
- (i2d 2f 2m 2 P, +
(8.45)
kl
012
(8.46)
021
(-_p2 +i2 2 0
-2 )q2-(i 2 dlfllp,
+f2ml,)Udl021
c1
(i2~d 2 f mI
2 2 P,
2-2 )-d2P
+f2 m2)d2
2
ki
k,
(8.47)
Ifp, = P 2 ,
(-p
2
1
+ i2
O) P, 2 2+ 2 q--(i2dfmlP
COI
CO(
+ fml
+
f2 -))Udl- 2 (i2d 2f2m 2pI +f 2 m2)d2
021
3k,
(8.48)
+ 021,q2).-
(8.49)
Equation 8.49 can be further rewritten by working with the solution expressed in terms of
complex quantities and using the notations and definitions presented by Equations from
8.32 to 8.37
M (0, , I q
q + 021 q,
Ml Ip+i2rdlflmi, +f, -Ml)U d -- m
(8.50)
In the same way, Equation 8.4 can also be rewritten by incorporating Equation 8.7
md2Ud2 + C/2Ud
2 kd
d2 (0-md12
+ 22q2)
(8.51)
(-m2 2
120
+i2df2m2Pl
+ f 22
2
22)Ud
M2(0,12 2P
1 , +022 q2) = -
(8.52)
If the stiffness of the primary structure is configured to produce the first mode shape of a
linear profile, the solution of the displacement of node 2 due to a periodic excitation can
be expressed as
u 2 =0 12ql +0 22q2 = q -0.5q 2 = -H
-0.5Hq2
k
ei = He".
k
(8.53)
H = Hql -0.5H,2.
(8.54)
node 2 regarding the first mode excitation is almost the same as that obtained from the
approximate solution presented in the previous section, demonstrating the accuracy of the
approximate solution.
As illustrated in Figure 8.4, both TMDs were tuned to the first mode. Thus, the second
mode response was not controlled. In terms of displacement, this was still acceptable for
both node 1 and node 2. However, with regard to acceleration, this resulted in the
maximum acceleration occurring in the second mode resonance condition. The
acceleration at node 1 in the second mode resonance condition was greater than that at
node 2 in the first mode resonance condition (Figure 8.5). This phenomenon occurs
when the primary structure damping ratio is very small, such as 0.5% in this example. In
H = p 2H
(8.55)
121
0.5
1.5
rho
2.5
0.5
40
3D
20
10
0.5
1.5
rho
2.5
122
10
B
6
4
2
0.5
1.5
rho
10
8
6
4
2
o
-2
-4
HAl
12
2.5
Figure 8.5
123
HI plot. U 1=(P/K) HI
~11 = 0.5, ~12 = 1.0
~21 = 1.0, ~22 = -0.5
(01=1
(02 = 2.5
Mass Ratio M1 = 0.003
Mass Ratio M2 = 0.030
Struc. Damping Ratio ~I = 0.005, ~2 = 0.0132
fl = 2.50~ f2 = 0.97
TMD Damping Ratio ~dl = 0.03, ~d2= 0.11
3.5
3
2.5
2
1.5
0.5
0
0.5
0.5
'1
II
3D
I\
25
2D
I\
15
10
5
)
..-/
0.5
\,
....,'---------.
1.5
rho
2.5
Figure 8.6
124
HAl
5
4
3
2
0.5
1.5
rho
2.5
1.5
rho
Figure 8.7
125
Md1
~j
;-
Md2
U1 + Ud1
d1
~\
. -
~ d2
U2+
1j
c-d(n-1
Ud2
Un-1+Ud(n-1)
Un + Udn
K2
K1
--_/\I\I\!\
(.
.. ,' -
-:;
Md(n-1)
M1
M2
--1
C1
~~U,
-
Figure 8.8:NDOFINTMD
P1
Mn-1
Mn
- - - - -----j
--:::G
C2
Cn-1
L:~~
ll~
Un-1Co
P(n-1)
L:u
Pn
System Model
(8.56)
(8.57)
(8.58)
(8.59)
(8.60)
TMDs located on the floors close to the ground are not as effective as the ones located at
the higher floors with regard to the firstmode. The model shown in Figure 8.8 can be
modified and the equations shown above can be adjusted accordingly. The 6DOF/4TMD
system study shown in the next section demonstrates this strategy.
126
Mooi
M46
i-k
4
Mkcd3
4
IMd
IMo2 i
_ Figure 8.9:
6DOF/4TMD
+C,,d
+ C5,,dS) + (kd 6 Ud6 +Cd6Uld6 )
4U,4) + 5(k,5Ud5
mi/+c/i+ku= p+ 013(kd3 Ud3 +Cd3Uld
13 )+ (kl 4 Ud
4
016
'6
16+
nild3ild3
+ ,3ld3
+ kd3u,3
-=
-01 d3U
16
md4id4 + c,,4U
+ kd 4 ud4
014
M4u
(8.61)
(8.62)
(8.63)
016
Mld Udsi+CC6ud
IudS
6
m15iid
nlld
6ii(1 6
+k
+ k15Ud5
1+ drld5
+Cd 6
md,
15
615
0
id16
+ kd16 Ud6 : -n-md6
(8.64)
(8.65)
Through the process presented in Section 8.2.1, the solution due to periodic excitation
can be expressed as
Pi
i =z He'
(8.66)
Here,
amplification factor.
The study is performed in four steps. First, only one TMD, the mass ratio of which is
0.01 (n,
mi,
(Figure 8.10). Second, the four TMDs are installed simultaneously, employing the same
tuning parameters used in the first step (Figure 8.11 a). Third, the four TMDs are
installed simultaneously, and tuned optimally for this new 6DOF/4TMD system (Figure
8.11 b). Fourth, at node 6, a single TMD, which is equivalent to the four TMD's at nodes
6, 5, 4, and 3, is installed, and tuned, and its response is compared with that of the
127
Md6\
a) TMD at Node 6
-I
Mos|
I-
Mo4 1
4 = 0.005
5d6= 0.065
f6 = 0.989
m6 =0.01
12
I'
8-
,\
./
4-
Mo
6-
Mo
2
-,
10-
0:9
.8
11
12
rho
b) TMD at Node 5
= 0.005
'/
14dS =
0.055
f5 = 0.992
m5 =0.01
Mo3
12-
i.
10
/
66-
Mo2
iMor
4.
18
0.9
1'1
1:2
11
rho
c) TMD at Node 4
Mo
6
=40.005
Mo
5d4 =0.045
Mo4 kd MdI
Mo3
f4 = 0.995
m4 =0.01
.i
i
i
i M1
Mn_
K3R
09
.2
rho
d) TMD at Node 3
= 0.005
MoI
Mo3jJ,
Mo2
MI1
iE
0.035
f3 = 0.997
m3 =0.01
4d3 =
Mos
Md
3
5IB
8
0A9_1
0.9
1
rho
12
11
128
12
M04 1
Mo3
Md4
k
kd31
~
cd3IMd3
d6 = 0.065
d5 = 0.055
5d4= 0.045
d3 = 0.035
f6 = 0.989
f5 = 0.992
m6 =
m5 =
m4 =
m3 =
Moi
10-
/
8-
f4 = 0.995
f3 = 0.997
Mo2
12-
I/
0.01
0.01
0.01
0.01
I/
',
I/
\x
'I."/
''I~~~~~'
i ,~~~~~~~~,
38
0.9
i
rho
1.2
1.1
M 3:-I
Md3
Mo2
Mo
7_
d6 = 0.11
4d5 = 0.10
5d4 = 0.09
d3 = 0.08
f6 = 0.961
f5 = 0.978
f4 = 0.985
f3 = 0.990
m6 = 0.01
m5 = 0.01
m4 = 0.01
m3 = 0.01
B8
I
I~~~~~
.
0.e
0.9
1
rho
~~~.
.
1.1
1.2
Mo3
Mo2
Moi
JMd
4d6 = 0.11
4d5 = 0.10
d4 = 0.09
4d3 = 0.08
f6 = 0.961
f5 = 0.978
f4 = 0.985
f3 = 0.990
m6 = 0.024
8-
7-
/' /
/~
/1
/'
43-
/
_
0.8
0.'9
121
rho
129
As can be seen from Figure 8.10, if a single TMD is installed at different nodes, the one
at the top of node 6 is most effective in reducing the first mode dynamic response of the
system with stiffness calibrated to produce the first mode shape of a linear profile. As its
location is lowered, its equivalent mass ratio and, therefore, its effectiveness decreases.
Also, even though the TMD mass of each case in Figure 8.10 is the same, its optimal
tuning parameters are different depending on its vertical location.
Figure 8.11 (a) shows the plot of the dynamic amplification factor of the system that
combines all the four cases shown in Figure 8.10. As can be seen from the figure, merely
combining the four optimally tuned cases as they are does not produce optimal behavior
in the combined new system. The system is further tuned for optimal behavior and the
result is shown in Figure 8.11 (b). As can be noticed by comparing Figures 8.11 (a) and
mi
, =I2I
McI
(n = 1, 2, ... 6) .
(8.67)
The mass ratio is defined in terms of the TMD mass and the equivalent mass of the
primary structure:
(8.68)
m-=
rle
mn,)
For the first mode, DInis 1 for node 6, and it gradually reduces as node number decreases
toward the ground.
m,
mass ratio m gradually decreases as node number decreases because ii,,e increases.
For the 6DOF system, the first mode shape of which is a linear profile, if a single TMD is
installed at node 6, this configuration requires a total of 40% less mass compared with the
distributed system shown in Figure 8.11 to achieve the same equivalent damping ratio
(Figure 8.12).
130
x 100%
40%
(8.67)
DAFIiV,,,
Since
by
(8.68)
/I-v2
DAFdisp,
'
2J
(8.69)
Expressing the allowable maximum dynamic amplification factor Hi,,P,.for the structure
with TMDs in a similar form provides a measure of the necessary equivalent damping
ratio , with regard to displacement:
1
2Hlipl.
,,
(8.70)
A similar procedure is repeated for the desired equivalent damping ratio , with regard to
acceleration. The allowable maximum acceleration at the top of the primary structure is
established. On the basis of this acceleration constraint, the allowable maximum
dynamic amplification factor of the primary structure H,e, is determined. Then, the
necessary equivalent damping ratio Xcwith regard to acceleration can be expressed as
2H,
-- =--
(8.71)
In order to satisfy both the displacement and acceleration requirements, the equivalent
damping ratios estimated using Equations 8.70 and 8.71 are compared, and a higher value
is selected as a design parameter. Once the desired equivalent damping ratio is obtained,
the mass ratio ( a, / m ) based on a SDOF/TMD system is determined from Figure 8.13.
131
'_ .......
(>.18
......
_. ,
'
/-. ....................
: .................. ..
0.(J4
_(>.,,
-........
. .; . . ...........
..............................
-
0.il2
...............
...---
= 0
0.02
f,
0
........
.........
0.02
0.03
0.04
.........
005
0.06
0.07
0.08
0.0
0.1
Since we distribute TMDs vertically, not every TMD mass is as effective as the ones at
the top. Further expanding Equation 8.67, the i'h mode equivalent generalized mass of
the primary structure that has a TMD at its node n is
]M = 1
mnie
}iM
(8.72)
The mass ratio is defined in terms of the TMD mass and the equivalent mass of the
primary structure.
= md
(8.73)
mie
As was discussed in the previous section, if we distribute TMDs vertically, we need more
mass overall, compared to the conventional huge single TMD scheme, in order to achieve
the same equivalent damping ratio. The additional mass required to compensate for the
loss of effectiveness due to vertical distribution can be estimated based on the mode
shape and the vertical range of distribution from the top to a certain level for the first
mode. If the stiffness of the structure is calibrated to produce the first mode shape of a
linear profile and TMD masses are evenly distributed for every node, the total additional
mass required is increased as the number of nodes that have TMDs increases. However,
the advantage of distributing TMDs over a broader vertical range is that each TMD mass
becomes smaller as the range increases to achieve the same equivalent damping. In
addition, due to their small sizes, the installation of TMD can be much easier. Thus, the
For a 60-story building whose first mode shape is a linear profile, for example, if TMDs
are evenly distributed for the upper half of the building from the 60t l to the 3 1stlevel, this
configuration requires a total of 67% more mass compared with the conventional system
that has a single TMD only at the 6 0 th floor.
-1
x100%z67%
(8.74)
However, compared with the conventional system that has a single TMD at the 60t h floor,
only 5.6% TMD mass is required for each level by this distribution. Furthermore, in this
study, the total required TMD mass for each level represents several dozen small TMDs
installed within the DSF system cavity space. If the TMD mass for each level is
distributed
to 60 TMDs, for example, each TMD mass is less than 0.1% of the
conventional huge single TMD located near the top of the building.
Once TMD mass ratio m for each level is determined, every TMD at each level can be
preliminarily tuned using that mass ratio. The optimal frequency ratio f, can be
preliminarily estimated using Figure 8.14. Then, preliminary TMD frequency codand
stiffness k can be computed as follows:
0ro= f r,
(8.75)
k, = m,,Co,
(8.76)
111
133
The optimal TMD damping ratio a, ,p, can be preliminarily estimated using Figure 8.15.
Then, preliminary TMD damping parameter c, can be computed as follows:
C = 2rd
(8.77)
lop,. amd
1
71-1
As was discussed earlier, tuning every TMD for each level based on its mass ratio for
each level does not lead to optimal performance of the combined final system. Thus,
these processes are used for preliminary tuning only. Iterations are necessary for optimal
tuning for the combined new system.
134
C1 +C2
+CJ1
-C2
o
o
o
o
-C2
C2 +C3
+CJ2
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
- CJ(n-l)
o
o
o
kn_'
+ kn + kJ(n_,)
-Cn
kn +CJn
o
o
k,/(n_')
o
o
o
o
UJ(n_l)
o
o
o
o
o
o
o
CJ(n-l)
o
o
-kn
o
o
-C'/(n_l)
o
o
o
o
o
o
o
o
o
o
o
o
-kJ(n_l)
o
o
o
o
o
o
k,/(n_')
Pn-I
Un
Pn
o
o
o
o
o
(8.79)
When TMDs are not installed to the nodes close to the ground, the matrices shown above
are reduced accordingly.
135
70 MPH
---~
----JI
L
I
1~
--=-_=J
[-
~~
=:J
I
-~
~I
-----<I
-1
__
J
Figure 21
13.7 PSF
~~~JI~
19.6 PSF ~
9.8 PSF
13.7 PSF
t ( VORTEX
SHEDDING
FREQUENCY)
022 HZ
,
( SlOG
0.18 HZ
FREQUENCY
)
- - - - - - - - . - - - - - - - - - - ~K-t4
I
I
I
L__ .
__
70 MPH
136
~_
100 MPH
V(wtlDVELOCllY)
0.4
- -- -.- -- ~_...
0.3
:-
: .......
_.0 - - .:. - - - .
_.~ ..
] ....
:.............
-. ~- - - -- - --.
t- ---- ~
--- ._~_.......
i!
..
,
,
; ......
0.35
.2
..
..
-:. - - -. - -. -
:::::
j o~ .::::::T:::::::L::::::::::::::::T:::::::J:::::::::: ::::::::
~
E
0.15
; .. -------..:
----- .. :
r r".
,
0.1
~.1~~ t ~ !
0.05
:.Q.8'
:.Q.6
:.Q.4.
:.Q.2
0
0.2
amplrtude
0.4
06
0.8
---
structure
o.
--
..
---
------
__
__
!!!!!!!!!!!I!i!i!!!!!iI.!il!i!!!!
iIIlili!iiiiii!!!!!!i!!!iilil!!ii!!
1
-----
1"
--------
..
no
structure
..
..
a_a
0.5
..:
-r
3
4
modal period
--.-----------------------..............................
- --- .. --_
- ---- ...
-------- -------- ---,------- -----,. - ---------- ------.--, - .. ------....... --,
- ,.
, - .. ----------r
-,.
.. .
..
--;----
1.5
lilllllil'lli
.... -. - - - - - - - - - - --.-r'--- - - - - - __ - - - - - -. - - - - - - - .- .
.......-----,..
-.-----------..._
--or -. --------
- -,- --
_ --r
-
--
--
---
,.
---,-
_0'
_0
-
-- - --- --
0_
:::::::::::::1:::::
__
,..
,
, - --- - -
r -
r - ---
, -
,- --- --- -
,.
-r-- -- ...
~m~~~~~~~~~m~~~~~~~~~~~:~~~~n~~~~~U~~~~~~~~~~H~~~~~~~~~~~~~
...............
-. ------ - ----
--
.........
- .-
--- - --r_
--
-.r--.
,- ----r - .............
, ,. - - -- -----., ,
- --,.... _------. --- ,-- - .. ------,.--------,
-... -------
::::::::::-
..
.:1:::::::::::
_--.- - ----
..
----- --- - -- ..
,..-.----.
,.._ .
r -- -
r
,.,..-
;~j:;~~~~~~~m~~~~~~~~~~~~~~~~~~~~~~~~~~~1~~~~~~~~~~~~~~~~~~~~~~~m~
0.05
0.1
0.15
0.2
Figure 8.18
Figure 8.18 contains first through fourth mode shapes of the structure, modal damping
ratios, maximum nodal relative displacement, and maximum nodal total acceleration. As
can be seen from the figure, the original structure, which has a very low inherent first
mode damping ratio (0.5%), does not meet the general acceleration requirement, 0.02g.
In resonance condition, maximum acceleration at the top is about 0.3g, almost 15 times
that of the acceptable value. In addition, maximum displacement at the top is almost 2.4
m (H/1 00), which is about 5 times the generally employed maximum displacement
parameter in practice, H/500.
In the following sections, both the conventional and
distributed TMD design strategies are investigated to satisfy both the acceleration and
displacement design criteria. The study showed that the contribution of the third mode
was negligible. Thus, only the first and second mode response controls are presented.
137
i2 m 1
n2
k=
24,603333Kg
602
12(24,603,333) = 24,603,333N /m
p = 0p = 610,000N.
The maximum displacement and acceleration parameters u*and a*are given by
u
500
= 0.48m
0.48
610,000 )
PHk 24,603,333
1
2H,
= 0.026.
2(19.4)
(pI(
~L2 2H
are
0.197
a
2
610,000
24,603,333
- 2(7.9)
=0.063.
Comparison between
:,Land ~e2 clearly shows that the acceleration controls the design
138
On the basis of the second mode shape and frequency of the system obtained from the
MotionLab eigenvalue analysis, the mass, stiffness, and force of the equivalent SDOF
system is estimated as
17
= 0 Mq = 14,866,870Kg
k=
p
=-2.52(14,866,870) = 92,918,000N / m
p = 185,400N .
--- = 0.48m
500
a* = 0.02g = 0.196m/sec 2
The dynamic amplification factor with regard to displacement H and the consequent
equivalent damping ratio
HI
0.48
0.48
=241
( 185,400
( 92,918,000)
1
c are
2H
2(241)
0.002.
The dynamic amplification factor with regard to displacement H2 and the consequent
equivalent damping ratio 5,2 are
0.197
2-
2H 2
14,866,870
e2
p 8 5 ,4 0 0
1
-
2(15.8)
0.032
Comparison between , and 4,2 clearly shows that the acceleration controls the design
with regard to the second mode response.
139
Scheme
In conventional TMD design, a huge single TMD is located near the top of the structure.
In order to achieve the 6.3 % first mode equivalent damping ratio estimated in Section
7.4.1, about 2.50/0 mass ratio is required (md = 615,000 kg: 0.850/0 of total building mass).
With this n1ass ratio the TMD parameters for optimal tuning are: 1) TMD Stiffness:
593,000 N/m (f= 0.98) and 2) TMD Damping: 135,000 Ns/m (11 %).
Forcing Period: 6.8 sec (First Mode Period of the Structure)
Modal damping ratio without feedback
oo~ ::::::::
o
--------
0.3
02
0.15
:g
0.1 ..
0.(5
~..
L::::::::t:::::::::::::::::::]::::::::: :::::::::::::::::::::
..--------
.. --------------
+]~ ~ ~ ~ [
.. --------------.--.
.
g> 0.25
.~
~
t ~
- -- --.............
.....
..
__
---
~,.,. ..
-- ..,- -------
__
--
::::: ::::::::J:::::::::::::
..
structure
-. - .. -- ,.. -- -- -- - - -- - .. - ,.
-,
,.
--..
.
-".
.:::::::::
,.
:::: ]:::::::::::::
..
::::::::::::: i::::::::::::::~:::::::::::::i::::::::::::::~::::::::::::
..........................
_____________
~
J
I..
__
------------- -----..........
--- .. ----.----------------------- -------- .. -- ---------.....
-- - ---- ..----- .. ----- ----- - ----~......... --.- ,.---- .. ----------r---------
J
J
__
..
..-
J
J
-.-
.., -.-----
--.------
0.05
....
.............
--.-
__
..
_
__
-.--
-.
,..
--.------ .
-.-----r-.---.-
~~~~n~;
~i~~
~mHH~m~mm m~m~~~~i~
~~~~~~H~~~~I.__________
__
~
J
J
__
~~~~~~~~~~~
I.
.1.
__ __
..
.
0.1
0.15
max nodal relative displacemenl
-OBO
50
100
150
200
250
:DJ
time in 98conds
350
400
450
500
200
210
215
time in seconds
220
225
Figure 8.19
140
Figure 8.19 shows the first through fourth mode shapes of the 60DOF/TMD system,
modal damping ratio, maximum nodal relative displacement, maximum nodal total
acceleration, and the time history of the nodal and damper displacement. By installing
the TMD at the 60th node, the damping ratio increases from an initial 0.5% to around 60/0.
Both maximum displacement and acceleration are lower than the target maximum values.
The nodal and damper displacement plot shows that they are about 90 degrees out of
phase, which is close to the theoretical optimum condition.
Forcing Period: 2.5 sec. (Original 2nd Mode Period of the Structure)
Profile of the maMimum nodal relative di6placement.primary
structure
nodaltolal
accelerutionlg-primary
structure
Figure 8.20
As can be seen from Figure 8.20, when the TMD is tuned only for the first mode, if the
building's second mode is primarily excited, the structure does not meet the acceleration
criteria.41 In the second mode resonance condition, the acceleration at the top is about
0.033g, which is much larger than the acceptable maximum value, 0.02g.
In order to
achieve the 3.2 % first mode equivalent damping ratio estimated in Section 5.1.2, about a
0.5 % mass ratio is required (md = 74,300 kg: 0.1 % of the total building mass). With this
mass ratio the TMD parameters for optimal tuning are 1) TMD Stiffness: 451,500 N/m (f
= 0.99) and 2) TMD Damping: 165,000 Ns/m (110/0).
This is only true when the inherent damping ratio of the primary structure is very low, such as 0.5% in
this study.
41
141
O~~
.~
0.3
g' 0.25
.~
~
:g
0.2
0.15
0.1
...... _- .. ---
------- . -- --.
-0.4
0.2
ampl~ude
-0.2
0.4
0.6
----
....
...
-----'-------
..
... ---------
.. ---
-'----------
--------
i---..--:.
---- -.
f -- .. -- . -- i--.
-- --:.
-- -..--
1~--.~.--t--~'~"'i--"""':-"""" t
0.05
-0.6
'::::::T:::::::1:::::::::[:::::::::::::::::::l::::::::: ::::::::
-OB
_._~---------.-----
--.
00
O.B
structure
nodal displacement
damper displacement relative to node
-0.1
-0.3
-0.40
10
20
40
50
60
time in seconds
70
eo
00
100
60
61
62
63
time in seconds
Figure 8.21
As can be seen from Figure 8.21, with a 0.5 % mass ratio, the second mode is controlled,
and thus the maximum acceleration is lower than the target maximum values. The nodal
and damper displacement plot shows that they are about 90 degrees out of phase, which is
close to the theoretical optimum condition.
142
6OIIt~.
TMDSARE
TUNED TO
T HE 2ND MODE
I'~-:
I'~-'
--:
SOthFL.
..---:
40thFL.
TMDSARE
TUNED TO
THE 1ST MODE
.~:
3<llh"
20thFL
10thFL.
-I .0
-0.8
-0.6
-0.4
-0.2
0.2
0.4
0.6
0.8
..-. .
::-.
:--.
T MDSARE
T UNEDTO
THE 2ND MODE
~-
1.
Simulation
143
For the second mode vibration control, we distribute TMDs from node 30 to node 21.
This configuration
However, this
configuration prevents TMD congestion near the top floors. It is a design decision how
to distribute multiple TMDs vertically. Table 8.1 shows TMD tuning properties at each
node. Md at each node represents the sum of possibly several dozen small TMDs.
Table 8.1: Vertically Distributed TMDs Tuning Properties
144
Node
Md (Kg)
60
32800
30862
0.9700
7318
0.115
1st Mode
59
58
57
56
55
54
53
32800
32800
32800
32800
32800
32800
32800
30893
30925
30957
30989
31021
31053
31085
0.9705
0.9710
0.9715
0.9720
0.9725
7306
7293
7281
7269
7257
0.115
0.115
0.114
0.114
0.114
1st Mode
1st Mode
1stMode
1stMode
1stMode
0.9730
0.9735
7245
7232
0.114
0.113
1stMode
1stMode
Kd (N/m)
Cd (N-s/m)
id
Tuned to
52
32800
31117
0.9740
7220
0.113
1stMode
51
50
49
48
47
46
45
44
43
42
41
40
39
38
37
36
35
34
33
32800
32800
32800
32800
32800
32800
32800
32800
32800
32800
32800
32800
32800
32800
32800
32800
32800
32800
32800
31149
31181
31212
31244
31277
31309
31341
31373
31405
31437
31469
31501
31533
31565
31598
31630
31662
31694
31727
0.9745
0.9750
0.9755
0.9760
0.9765
0.9770
0.9775
0.9780
0.9785
0.9790
0.9795
0.9800
0.9805
0.9810
0.9815
0.9820
0.9825
0.9830
0.9835
7208
7196
7183
7171
7159
7146
7134
7121
7109
7097
7084
7072
7059
7047
7034
7022
7009
6997
6984
0.113
0.113
0.112
0.112
0.112
0.112
0.111
0.111
0.111
0.111
0.110
0.110
0.110
0.110
0.109
0.109
0.109
0.109
0.108
1stMode
1stMode
1stMode
1stMode
1stMode
1stMode
1stMode
1stMode
1stMode
1st Mode
1st Mode
1st Mode
1st Mode
1stMode
1stMode
1stMode
1stMode
1stMode
1stMode
32
32800
31759
0.9840
6971
0.108
1st Mode
31
32800
31791
0.9845
6959
0.108
30
29
28
27
26
25
24
23
22
21
25300
25300
25300
25300
25300
25300
25300
25300
25300
25300
153417
153417
153417
153417
153417
153417
153417
153417
153417
153417
0.9850
0.9850
0.9850
0.9850
0.9850
0.9850
0.9850
0.9850
0.9850
0.9850
6958
6958
6958
6958
6958
6958
6958
6958
6958
6958
0.055
0.055
0.055
0.055
0.055
0.055
0.055
0.055
0.055
0.055
1st Mode
2nd
2nd
2nd
2nd
2nd
2nd
2nd
2nd
2nd
2nd
Mode
Mode
Mode
Mode
Mode
Mode
Mode
Mode
Mode
Mode
04
0.35
O.B
...
.
.
....................................
--
.
,
..........
L.
__
,
I
--..
.
..
,:
',:
nodal displacement
- .. -
,
I
:::J:::r::::rll:Iu:
--
i.
"
- .. -----,.-
i.
- .. r
-.--.-------.--.-I
..
.i ~
r
--
..
t..m m.t ~.
.(J.B
20
,(0
9.l
00
100 120
time in seconds
140
100.
180
200
Figure 8.23: Dynamic Response of the 60-Story Structure with Vertically Distributed TMDs tuned to
the First and Second Modes when Forcing Period = First Mode Frequency of the Structure
Figure 8.23 shows modal damping ratios, maximum nodal relative displacement,
maximum nodal total acceleration, time history of the nodal and damper displacement,
and the maximum inter-nodal displacement of TMDs tuned to the first mode in the first
It can be noticed that the maximum nodal relative
mode resonance condition.
displacement and nodal total acceleration are almost the same as in the case with a single
TMD at node 60 in Section 8.4.3.
145
0.15
0.4
o:::O::1.-.:o:L::
0.1
!:o~-:.:.:, :.:}[:.:.:
0.05
1 :: .:: }:
0.1
0.05
~..
~J~ ~
t~
i: :
t ~ .
t t ~
3
4
modal period
Q>
1..,
o
c
__,: .::::.::.::'::::::".'::::::1:~':::::.::.':::.:
Q>
1..,
Q>
6 . --
--
o
c
4 ....
00
oo!
u.oo.!
----r
---
-----
00
1
0.134
,,
~!
;.
00
\"""
4-
.00.
00
..
__ j
..........r"'----.r
..
--
i"
00
00
i'"
00
00
~u
00
\"""
--.
00
--
'
0.136
Figure 8.24: Dynamic Response of the 60-Story Structure with Vertically Distributed TMDs tuned to
the First and Second Modes when Forcing Period = Second Mode Frequency of the Structure
Figure 7.24 shows modal damping ratios, maximum nodal relative displacement,
maximum nodal total acceleration, time history of the nodal and damper displacement,
and the maximum inter-nodal displacement of TMDs tuned to the second mode in the 2nd
mode resonance condition.
It can be noticed that the maximum nodal relative
displacement and nodal total acceleration are almost the same as in the case with a single
TMD at node 60 in Section 8.4.3.
146
The motion of the TMD mass often acts as a limiting factor of TMD design because
TMDs are installed within the limited interior space of a building. Figure 8.25 illustrates
the ratio of maximum TMD displacement to maximum primary structure displacement.
As mass ratio (TMD Mass/ Primary Structure Mass) decreases, for optimal performance,
the motion of TMD must become greater, requiring more space to accommodate this
motion. This space requirement due to structural requirements may conflict with that due
to architectural or other requirements.
Ud/U
1
I1
In the distributed TMD system presented in this chapter, multiple small TMDs are
installed within the DSF cavity space, the depth of which typically ranges from 0.3 to 1.5
meters (Arons, 2000).
When the TMDs are designed to move in the direction
perpendicular to the facade plane, the motion of TMDs should be accommodated within
the depth of the DSF cavity, or the depth of the DSF cavity must be wide enough to allow
TMD motion for optimal performance. This TMD-motion-induced space requirement
can be reduced by decreasing TMD motion. One way to reduce TMD motion is to
increase TMD damping ratio over optimal value. This strategy is examined using the
distributed TMD model presented in the previous section. The TMDs tuned to the
second mode are omitted in this study because they move much less than those tuned to
the first mode.
147
When the first mode of the primary structure is excited, the motions of the TMDs tuned
to the 1st mode range from 0.4 to 0.75 meters, and the TMD damping ratios from
10.8% to 11.5% for optimal performance. Suppose that the motion of the TMDs should
be limited to maximum 0.5 meters due to the limitation of the DSF cavity space. This
limited motion can be achieved by increasing the TMD damping ratios. Table 8.2
contains new properties of the TMDs employed in the previous section to meet this TMD
motion constraint. Tuning was performed by iteration. As can be seen from the table,
now the TMD damping ratios range from 15.5% to 29.5%, which is much higher than the
previous case.
Tuned to
Md (Kg)
Kd (N/m)
60
59
58
57
32800
32800
32800
32800
30862
30893
30925
30957
0.9700
0.9705
0.9710
0.9715
18795
18655
18450
18202
0.295
0.293
0.290
0.286
1st Mode
1st Mode
1st Mode
1st Mode
56
55
54
53
32800
32800
32800
32800
30989
31021
31053
31085
0.9720
0.9725
0.9730
0.9735
17932
17646
17350
17044
0.281
0.277
0.272
0.267
1st Mode
1st Mode
1st Mode
1stMode
52
51
50
49
48
47
46
45
44
43
42
41
40
39
38
37
32800
32800
32800
32800
31117
31149
31181
31212
0.9740
0.9745
0.9750
0.9755
16730
16410
16088
15763
0.262
0.257
0.252
0.246
1stMode
1stMode
1stMode
1st Mode
32800
32800
32800
32800
32800
32800
32800
32800
32800
32800
32800
32800
31244
31277
31309
31341
31373
31405
31437
31469
31501
31533
31565
31598
0.9760
0.9765
0.9770
0.9775
0.9780
0.9785
0.9790
0.9795
0.9800
0.9805
0.9810
0.9815
15435
15106
14777
14448
14120
13790
13460
13130
12799
12468
12138
11820
0.241
0.236
0.231
0.225
0.220
0.215
0.210
0.204
0.199
0.194
0.189
0.184
1st Mode
1st Mode
1stMode
1st Mode
1stMode
1stMode
1stMode
1stMode
1stMode
1stMode
1stMode
1stMode
36
32800
31630
0.9820
11500
0.179
1st Mode
35
34
33
32
32800
32800
32800
32800
31662
31694
31727
31759
0.9825
0.9830
0.9835
0.9840
11200
10900
10600
10300
0.174
0.169
0.164
0.160
1stMode
1stMode
1stMode
1stMode
31
32800
31791
0.9845
10000
0.155
1st Mode
148
Node
inter-nodal
displacement
dampar
I::::~:::m:::m:H'~H.+.,~."T.i
.::.:
.1:1:~:~~::~:!::::~::~~~~!:~.::.::~:::.~
~.~~::j~:::.~t::.::~:
llllllllllll.lllllllllllllllll11111l llll
m
:492
0.494
0.496
0.498
, l11axim~m;i~!er-nodal
displacement
1111.1~
0.5
0.502
i1: ::!'Lilt.i~';.'~fi\ili:l!j;l:il~]l:!l~i.~/I:I.
-05
;.;
20
40
i ;:1
I : ;
00
..
i . , :
00
100
120
time in seconds
! .
140
1 ..
'j ..
Ifjll
HIll
200
structure
__
..
__
..
~~~m~mm~m~n~~H~~~m~~~~~i~~~~~~~~~~~i~~~~~~:;;i~i~~~~~~~~~~
_. __
_.
._L
L_
L
L
-.-
......
----------...
..........
--- ---- ---- ...
........
--- - - -------.-.----........
-- ,.-.- - -.-- ----- ---...
-- -- ,
-....
, -- ,,.
..................
,
............ ,
,-- .._
._
_
__
--.------- .
-- ----- ..
--.---..
---.
--- -.-- -.- ----..
,, ,,.. -,,
,,
...
--
..
~;;~~~~~~~~~~~~~~~~~~~~~~~~~~~:~;:~l~~~~~~~~~~~l~~~
.......
_._
__
L
L _
"
__
L_
........
--- -.----.......
.. -- ----- - .. - ---- --........
.. --.- -.-- ------ -----.-....
.. - ---.- -- .. - --- -.--, .. - --- .. ,--- -.-- .. ,---.-...----,.--........ -- -.- ,
-.--.,
-.-,
,
, --,
_
_._
__
- .-- ..
-- .
- -..
- .-
..
mm
~~n~~~~mmm mm~m
I ~~~~~~~~~m ~~~~~~~HI ~~~~~~~~~..
........
___
---
_L
__ .L
_L.
_. __ .. _.L
0.005
__
__
001
ll.D15
0.02
max nodal tolal acceleration I 9
_ .. _
_
..
0.025
Figure 8.25: Dynamic Response of the 60-Story Structure with Vertically Distributed TMDs with
Higher Damping Ratio for Reduced TMD Motion
Figure 8.25 contains the maximum inter-nodal displacement of TMDs, time history of the
nodal and damper displacement, maximum nodal relative displacement, and the
maximum nodal total acceleration of the system having the properties shown in Table 8.2.
As can be seen from the figure, by increasing TMD damping ratios over optimal values,
the motions of the TMDs are limited to the target value, :t 0.5 meters. However, the
system does not produce optimal performance, even though the TMD motion constraint
requirement is met. Now, the maximum acceleration is about O.028g, which is greater
than the generally accepted maximum target value, O.02g. Thus, in order to resolve this
acceleration problem caused by the strategy used to meet the TMD motion constraint,
another strategy must be considered. For example, increasing TMD Inass will resolve this
problem, if it is acceptable. Design decisions should be made in an integrative way
considering various complex and sometimes conflicting requirements.
149
--
~ --
--
_H
II
1) Conventional
150
huge TMD
at the top
(Alt.I)
rI
--
~
n
.~
TI
n
II
I
--~.
I
I
---,,-a.
I
I
--
-~.
--~
-~.
(Alt. 2)
(Alt. 3)
151
I'f-L. ~+
+ + + +
t t t t
+ + + + + +
t t t
+ +
t t
.-1
I
I
I
-.
:j
-:
.1
:,
I
I
II
I)
2)
III
/
I
I
3)
4)
-+--~.-+---"--+-_-+--.--i
152
t~
8.6. Summary
1) By vertically distributing TMDs, no building area is occupied by conventional huge
TMDs, which results in valuable space saving near the tops of tall buildings.
2) By distributing TMDs vertically, not only the first mode but
also other modes can be effectively controlled if necessary.
3) Since all TMD masses are located at the extreme perimeter
of the building, torsional resistance exerted by TMD is
maximized.
4) TMD installation becomes easy due to its reduced mass.
5) Vertically distributed small TMDs may be included within
the spandrel portions of DSF construction units. In this case,
the construction of DSF facades and vertically distributed
TMD systems can be performed simultaneously.
6) Vertically distributed small TMDs may be installed in some
other locations of a tall building such as underneath floor slabs.
Ilnit
....
..
-+
153
For the accomplishment of these expanded design objectives for tall buildings, integrative
design approach among architecture-related disciplines is of critical importance.
Basically, this thesis investigated contemporary technologies and their architectural
implications, and further, tried to develop new technologies and study their architectural
potentials. The interrelationship between technology and architecture is not determinate,
but rather is very dynamic. And this dynamic direction is strongly dependant on the
interaction between architects and engineers.
Integrative design approach is desirable for the entire spectrum of architecture from
vernacular housing to tall buildings worldwide. Among them, tall buildings were
selected as a research focus in this thesis to study the interrelationship between
technology and architecture because tall buildings are, in a sense, the accumulation of the
most advanced technologies due to their very tallness, and their architectural expressions
have significant impact on any urban context in which they soar due to their scale. Thus,
it was hypothesized that the interaction between technology and architecture is more
dynamic and, at the same time, easier to trace in tall buildings than in any other building
types. This research was performed based on this hypothesis, and proved that this
hypothesis is valid.
There are future research needs for the further investigations of the dynamic
interrelationships between technology and architecture not only for tall buildings but also
for other building types. These investigations will lead to higher quality built
environments for humans in this technologically advanced and esthetically plural era.
154
CONCLUSION
Modern architecture was created based on modern technology and grew into a new
architectural style that was preferred internationally. Tall buildings, which emerged in
Chicago in the late 19th century even before various modem architectural movements in
Europe, are clearly forerunners of mainstream modern architecture of the 20 th century.
This new building type was created through the development of iron/steel braced frames
and curtainwall concepts. Today, in this intellectually and culturally pluralistic era,
architecture is still deeply rooted in these original technologies, which enabled early tall
buildings/modern architecture to prosper as a new style.
Due to the imperfectness of technological originality and the conflicting dual
characteristics of any technology - promising reliability and ephemeral nature - original
technologies are followed by refining technologies. These refining technologies may be
called remedial technologies in the sense that they remedy unforeseen problematic
aspects of original technologies and make them perform as intended or even better. Even
though remedial technologies are mainly in the domain of function/performance, just as
original technologies created new esthetic expressions and in turn a new architectural
style, remedial technologies have great architectural potential, and architects and
engineers have been exploring this potential. Indeed, through this exploration, the style
created by original technologies is entering its culminating phase.
On the basis of the recognition of the symbiotic relationship between building systems,
the prevalent strategy to develop remedial technologies to overcome contemporary
technological limitations is design integration, which corresponds to the coevolution
process in nature. Through integrative design approaches, synergistic effects can be
achieved that eventually lead to architecture of higher quality.
In tall buildings, which in a sense are the accumulation of the most advanced
architectural technologies, the significance of integrative design approachs is more
important than in any other building type to overcome or at least minimize contemporary
technological limitations. Design integration requires intimate collaboration between
architects and engineers, and this cooperation cannot be achieved without a mutual
understanding between the disciplines. From a broader view point, this thesis is devoted
to helping enhance this kind of professional environment.
Understanding the technology, science, and mathematics behind the behavior of building
systems is the responsibility of contemporary architects who want to create higher quality
architecture. Likewise, understanding fundamental architectural design principles is the
responsibility of engineers who want to achieve higher quality engineering products that
are incorporated into architecture. This mutual understanding can become the
potentiality of enhanced design integration.
In Part I of this thesis, the history of tall buildings was re-illuminated through the filter of
the interrelationship between the technology and architectural style of tall buildings.
Breakthrough technologies allowed the emergence of a new building type, tall buildings,
155
and eventually led to a new architectural style through the esthetic aspiration of architects
who wanted to transform technological products into their esthetic ideology. While this
new style at its culminating phase is still a mainstream design direction, many branch-out
trends have been prevalent in tall building design. These pluralistic design approaches of
architects accompany the technological evolutions enabled by the efforts of engineers.
In Part II of this thesis, current tall building design practices were investigated, again
through the filter of the interrelationship between technology and architectural esthetics
developed in Part I.
research then focused on investigating two specific structural technologies and their
architectural implications. Based on the recognition of the motion-induced human
discomfort problem as a serious design issue in tall buildings, diagrid structural systems
were investigated first for their enhanced lateral stiffness and esthetic potential. With
regard to obtaining more damping against the dynamic motion problem, solutions were
sought through the integrative design approach between structural systems and facade
systems.
Each study, on its own, makes contributions in particular design situations. It is expected
that the study of diagrid structures - the optimal angle study, the simple member sizing
methodology, and other topics discussed, such as architectural, constructability, and
urban contextual issues - will be very useful to both architects and engineers for
preliminary design. Based on these studies, structural and architectural decisions in the
early stages of design can be made in a more integrative and efficient way.
Through the study of low stiffness double skin facade connectors and vertically
distributed tuned mass dampers, a new direction for solving the motion problem of tall
buildings was introduced. The studies on the theory, preliminary design guidelines, and
architectural implications of the distributed tuned mass dampers will be useful for both
architects and engineers. Within today's globally prevalent architectural context, which
values energy efficient design approaches more than ever, integrating structural motion
control with double skin facade systems has great potential.
From a wider angle, this thesis is intended to contribute to enhancing mutual
understanding between architects and engineers. For the author, this work is an initial
step toward many potential research projects that will lead to a more desirable
professional environment for architecture-related disciplines.
156
APPENDICES
157
158
V = KAu,
'I j
\/
/.=tn
.
N--
=-j
\
\
t .............
...... %
... n..,.L
49
X t;
7/
NI/
......
= 12 '
Cos 2 0
<--Both Sides
)
Lb
7
+ B EA ,, ) sin3
KB =KB7(B2 Ec
2L
(2Lc.
\
/
V--
Same as Above
K7 = 1 2 (
Lh
CEos2o
...
---P
--
=6
d2L
sin3 0
--
---"
...
12EIc +4 12EI,.
h3(1+r)
h3(1+ 2r)
r=
L
h I,
159
160
Building Classification: Category III (Buildings and other structures that represent a
substantial hazard to human life in the event of failure)
Gf = 0.9251+1.7Qgi
1+1.7gIz
30ft
z = 480fi
I = c(33/z) 6
c = 0.3 <-Exposure B
I z = 0. 19
= 3.4
g) = g,
gR = 21n(3600n) +
0.577
n(3600n)
/21n(3600nl)
Q: Background Response
1
=X
0.63
+ 0.63 LB+h
L = B = 120ft.
h = 800ft.
= 320ft.
E=1/3
e-ExposureB
L 7 = 774
Q = 0.77
161
R= I -R,,RhRB(0.53+O.47RI)
, = Damping Ratio = 1% Assumed
7.47N
N, = (1+10.3N
)5 /3
Vz
Vz = b(-
(33)
(60)
b =0.45 a =1/4=0.25
(-ExposureB
N1 = 1.2
R,, =0.12
RI,
77
1 -e-2
2q72
R, =1 for
=0
Rh = 4.41
1 (1-e-2)
27q 2
RB =
R, =1
27 2
11
for
=0
RB =1
Ri = R, =1
12 (-e-2 ) for
for
=15.4L/V
>0
=0
Rl = 2.21
2 gR 2
l.2 Iz gCQ
+ gRRJ =
1+1.7gI
)
G = 0.9251 + 1.7IzVQ
162
0.82
CPIIW =
08
C71W=
05
j
Zg
- Exposure B
= 1200ft.
(- Exposure B
K z = 2.01
Velocity Pressure qz
q = 0.00256K,Kz,KV
2I
= 63.8psf
Design Wind Pressure at the top of the 240m-tall 60-story building is as follows:
pz = qGCp - q (GCPJ)
P:,it1 i,,(p
= 41.9psf
Following the same procedure, Design Wind Pressure at the mid-height of the 240m-tall
60-story building is as follows:
= 3 3.5 psf
PzI,,'(, )r,,id-,eigl,,
Pzlw(,,,id-heighl
21 .Opsf
163
Leeward Velocity
P::/II'@IO[J
Pressure
= 26.2p5f
Windward
~~~1.\',
\~~:::"'~~~" .
\~.-::::"'III
':
~).
'
Velocity Pressure
.'.::'.tlJ"
l~~:~~-=~-:;~
~
--~1'
!~!J0-ifi~,!
r,w.. ' .,) ..1."
t.f14'!'~"'~
t;., 1'111
.0.l, .. ,1,
'..... :'/
.~'.'.II.,,~'.~,..i~ .. "1"
~I.,
I.I~
I"".'
U~"." I ,,"
~..,..~,
41
b
-
r,y""
'I.t.- .
,.,.~
~",
I'
"
~- .r
~.~
~lo"
P::/II'@lIIid-helghl
= 21.0
p~l
,--roo
~.
P ::wlI'@mid-heighl
= 33.5
,.
1-
"
I"
I'
P::/II'@ha.\e
164
= 8.0 p5f
P ::lI'w@hu.I'e
= 12.6 psI
psI
~~~
1.39
1.08
0.85
0.64
0.61
0.53
0.55
0.51
O. gO
0.53
0.50
:l
2.16
0.52
'!i.
0.31
0.41
165
166
l=
J
h
..
1
EA
StraIn energy for a bar (A, L) IS -Fo = -0
2L
L tanS / 2
EAL
= --8
V=--8)
2. EAL
+--(82
4cosB
2
2
EAL
I
2
2
2
EAL.
+83)=--{(-v2cosB8)) +82 +83}=--8'
4cosB
l'
4cosB
EAL
V =
V__
EffectiveArea
= _4_c_o_s_B_
.,1' 8. = _E_A
__ 8 .,1' 8.
8
L2 tan B
2L sin B
(a)
2
The bar extensions are expressed in terms of the cartesian strains using
= T.8:
8.
(b)
2L sin B
l'
(T'
.1'.T
1
)8=-8
2
l'
D'."q8
",~
EA (T.,1'T.) = EA
2cos2 Bsin2 B
L sin B
L sin B [
0
2sin4 B
2cos
].
Bsin 2 B
167
(2 cos + 2 cos4
0).
D,eq
=3
1/3
1/3
EA 1/3
L
3 A
4 L
lcm.
Diagrid Model
L (m)
A (sq. m)
Teq(cm)
0.03
2.23
12
0.09
1.89
24
0.18
1.64
168
Teq (cm)
1.91
Appendix 5. Preliminary
of Empirical
~ ~5
& Verification
(H / B ~ 8)
with s = 1
V(KN)
M(KN-M)
1,325
3,975
6,625
9,275
11,925
14,417
16,820
19,223
21,359
22,961
24,563
26,165
1,988
31,800
127,200
286,200
508,800
794,526
1,140,267
1,543,947
2,004,498
2,515,512
3,066,576
3,656,088
Ad(bndQ)
0.0004
0.0060
Ad(Shear)
0.0241
0.0542
0.0964
0.1506
0.2161
0.2926
0.3799
0.4768
0.5813
0.6930
0.0148
0.0207
0.0266
0.0322
0.0375
0.0429
0.0477
0.0513
0.0548
0.0584
0.0030
0.0089
-+-Series1
___ Series2
0.8000
0.7000
0.6000
I!! 0.5000-
~ 0.4000
tT
0.3000 U)
0.2000
0.1000
0.0000
169
Story
67th - 69th
61 st - 66th
55th - 60th
49th - 54th
43rd - 48th
37th - 42nd
31st - 36th
25th - 30th
19th - 24th
13th - 18th
7th - 12th
1st - 6th
V(KN)
1,325
3,975
6,625
9,275
11,925
14,417
16,820
19,223
21,359
22,961
24,563
26,165
M(KN-M)
1,988
31,800
127,200
286,200
508,800
794,526
1,140,267
Ad(bndg)
0.0002
0.0036
0.0145
0.0325
0.0579
0.0904
Ad(Shear)
0.1297
0.1756
0.2280
0.2861
0.3488
0.4158
0.1126
0.1287
0.1430
1,543,947
2,004,498
2,515,512
3,066,576
3,656,088
0.4500
0.4000 0.3500 ~ 0.3000-
S
0.2500CD
E 0.2000C"
170
0.0089
0.0266
0.0444
0.0621
0.0799
0.0965
0.1538
0.1645
0.1752
-+-Series1
___ Series2
V(KN)
Story
67th - 69th
1,325
3,975
61st - 66th
55th - 60th
49th - 54th
43rd - 48th
37th - 42nd
31st - 36th
25th - 30th
19th - 24th
M(KN-M)
1,988
31,800
127,200
6,625
9,275
11,925
14,417
16,820
19,223
21,359
22,961
13th - 18th
7th - 12th
1st - 6th
24,563
26,165
286,200
508,800
794,526
1,140,267
1,543,947
2,004,498
2,515,512
3,066,576
3,656,088
Ad(bndq)
Ad(Shear)
0.0192
0.0577
0.0961
0.1346
0.1730
0.2092
0.2440
0.2789
0.3099
0.3331
0.3564
0.3796
0.0002
0.0033
0.0131
0.0294
0.0522
0.0816
0.1171
0.1585
0.2058
0.2583
0.3148
0.3754
~Series1
___
Series2
0.4000
0.3500 0.3000 ~ 0.2500-
rio~
~
<0()..~ <0~q. ~ ~~
rS;J~
~
~~
"
~Q)~
~~~
~,o 0;)'
A.~'
~o "",
~ fo~ ~ ~~
~q. l);;ro~ "Oj~
0;)
~~
~
~CO~
'\~
~~
,~
q.
"
stories
171
172
StOry
55th - 60th
49th - 54th
43rd -48th
37th - 42nd
31 st- 36th
25th - 30th
19th - 24th
13th - 18th
7th - 12th
1st- 6th
V(KN)
M(KN-M)
2,650
5,300
7,950
10,442
12,845
15,248
17,384
18,986
20,588
22,190
7,950
71,550
198,750
389,076
639,417
947,697
1,312,848
1,728,462
2,184,126
2,788,776
Ad(bndq)
0.0013
0.0118
0.0328
0.0641
0.1054
0.1562
0.2164
0.2849
0.3600
0.4596
Ad(Shear)
0.0059
0.0118
0.0177
0.0233
0.0287
0.0340
0.0388
0.0424
0.0460
0.0495
0.5000
0.4500 0.4000 0.3500 ~ 0.3000S
~ 0.2500
tT 0.2000 U)
0.1500
0.1000 0.0500 0.0000
173
Story
55th - 60th
49th - 54th
43rd - 48th
37th - 42nd
31st - 36th
25th - 30th
19th - 24th
13th - 18th
7th - 12th
1st - 6th
V(KN)
2,650
5,300
7,950
10,442
12,845
15,248
17,384
18,986
20,588
22,190
M(KN-M)
7,950
71,550
198,750
389,076
639,417
947,697
1,312,848
1,728,462
2,184,126
2,788,776
Ad(bndq)
Ad(Shear)
0.0008
0.0074
0.0205
0.0401
0.0659
0.0148
0.0296
0.0444
0.0583
0.0717
0.0976
0.1352
0.1781
0.2250
0.2873
0.0851
0.0970
S 0.2000 Q)
E
C"
In
0.1500
0.1000
0.0500
0.0000
~~
,<0
~
174
~~
9J~
~
,~
t1;)'o'
0.1059
0.1149
0.1238
--+-Ad
___
for Bending
Ad for Shear
Story
55th - 60th
49th - 54th
43rd - 48th
37th - 42nd
31st - 36th
25th - 30th
19th - 24th
13th - 18th
7th - 12th
1st - 6th
V(KN)
2,650
5,300
7,950
10,442
12,845
15,248
17,384
18,986
20,588
22,190
M(KN-M)
7,950
Ad(bndQ)
0.0007
0.0066
0.0182
0.0356
0.0585
0.0868
0.1202
Ad(Shear)
1,728,462
2,184,126
0.1583
0.2000
0.0296
0.0592
0.0887
0.1165
0.1434
0.1702
0.1940
0.2119
0.2298
2,788,776
0.2554
0.2477
71,550
198,750
389,076
639,417
947,697
1,312,848
-+-Ad
___
for Bending
Ad for Shear
0.3000
0.2500 -
0.2000 -
SCD
E 0.1500 C"
Uj
0.1000
0.0500 0.0000 -
175
...
a
1
5
s
176
/ B ~ 6)
Story
49th - 51th
43rd - 48th
37th - 42nd
31st - 36th
25th - 30th
19th - 24th
13th - 18th
7th - 12th
1st - 6th
V(KN)
1,325
3,975
6,467
8,870
11,273
13,409
15,011
16,613
18,215
M(KN-M)
1,988
31,800
126,726
281,667
494,547
764,298
1,084,512
1,444,776
1,954,026
Ad(bndQ)
0.0003
0.0045
Ad(Shear}
0.0178
0.0395
0.0693
0.1071
0.1519
0.2024
0.2738
0.0144
0.0030
0.0089
0.0198
0.0252
0.0299
0.0335
0.0371
0.0407
Ad for Shear
0.3000 0.2500 -
0.2000 -
SQ)
E 0.1500 C"
0
177
Story
49th - 51 th
43rd - 48th
37th - 42nd
31st - 36th
25th - 30th
19th - 24th
13th - 18th
7th - 12th
1st- 6th
V(KN)
M(KN-M)
1,325
3,975
6,467
8,870
11,273
13,409
15,011
16,613
18,215
1,988
31,800
126,726
281,667
494,547
764,298
1,084,512
1,444,776
1,954,026
Ad(bndg)
0.0002
0.0030
0.0118
0.0263
0.0462
Ad(Shear)
0.0714
0.1013
0.1349
0.1825
0.0599
0.0670
0.0742
0.0813
.e
~ 0.1000
C" 0.0800
U)
0.0600
0.0400
0.0200
0.0000
178
0.0059
0.0177
0.0289
0.0396
0.0503
-+-Ad forBending
___ Ad forShear
Story
49th - 51th
43rd - 48th
37th - 42nd
31st - 36th
25th - 30th
19th - 24th
13th - 18th
7th - 12th
1st - 6th
V(KN)
1.325
3.975
6,467
8,870
11,273
13,409
15,011
16,613
18,215
M(KN-M)
1,988
31,800
126,726
281,667
494,547
764,298
1,084,512
1,444,776
1,954,026
Ad(bndg)
0.0002
0.0025
0.0101
0.0225
0.0396
0.0612
0.0868
0.1157
0.1564
Ad(Shear)
0.0118
0.0355
0.0577
0.0792
0.1007
0.1197
0.1340
0.1483
0.1626
-+- Ad for
___
Bending
Ad for Shear
0.1800
0.1600 0.1400 0.1200 -
l!!
SQ) 0.1000-
E 0.0800C"
tn 0.0600-
0.0400
0.0200 0.0000 -
179
180
Story
37th - 42nd
31 st - 36th
25th - 30th
19th - 24th
13th - 18th
7th - 12th
1st - 6th
V(KN)
2,492
4,895
7,298
9,434
11,036
12,638
14,240
M(KN-M)
7,476
67,017
184,497
358,848
583,662
848,526
1,262,376
Ad(bndg)
0.0009
0.0077
Ad(Shear)
0.0213
0.0414
0.0673
0.0979
0.1456
0.0163
0.0211
0.0056
0.0109
0.0246
0.0282
0.0318
-+- Ad
___
for Bending
Ad for Shear
0.1600
0.1400
0.1200
~
0.1000
SQ)
E 0.0800 C0
181
Story
37th - 42nd
31st - 36th
25th - 30th
19th - 24th
V(KN)
13th - 18th
7th - 12th
1st - 6th
11,036
12,638
2,492
4,895
7,298
9,434
14,240
M(KN-M)
7,476
67,017
184,497
358,848
583,662
848,526
1,262,376
Ad(bndg)
0.0006
0.0058
0.0160
0.0311
Ad(Shear)
0.0083
0.0164
0.0244
0.0316
0.0505
0.0734
0.1092
0.0370
0.0423
0.0477
0.1200 0.1000 -.
f!
0.0800
S
Q)
E 0.0600
C"
tn
0.0400
0.0200
0.0000
182
Story
37th - 42nd
31 st - 36th
25th - 30th
19th - 24th
V(KN)
13th - 18th
7th - 12th
1st - 6th
11,036
12,638
2,492
4,895
7,298
9,434
14,240
M(KN-M)
7,476
67,017
184,497
358,848
583,662
848,526
1,262,376
Ad(bndq)
0.0005
0.0048
0.0133
0.0259
Ad(Shear)
0.0139
0.0273
0.0407
0.0526
0.0421
0.0612
0.0910
0.0616
0.0705
0.0795
Ad for Shear
0.1000
0.0900
0.0800
0.0700
~ 0.0600
S
~ 0.0500C" 0.0400 U)
0.0300 -
183
2500 -
2000 -
1500 c:
o
...,
1000 -
500 -
01
184
..:
u/H = 1/300
u/H = 1/250
185
u/H = 1/450
186
u/H = 1/400
H plot. U = (P/K) H
Mass Ratio (Md/M) = 0.01
Fayade Damping Ratio = 0.02, 0.04,
0.06, 0.08, 0.1, 0.2, 0.3
100
50
0.7
0.8
0.9
1
rho
1.1
1.2
1.3
2500
1.3
Hd Plot. Ud = (P/K) Hd
Mass Ratio (Md/M) = 0.01
Fayade Damping Ratio = 0.02, 0.04,
0.06, 0.08, 0.1, 0.2, 0.3
187
becomes minimum. Before passing this optimum damping ratio, as the damping ratio
increases, H (max) becomes lower. After passing this optimum damping ratio, as the
damping ratio increases, H (max) becomes higher. However, this H (max) increase is not
significant.
7.2. Secondary Structure (DSF Outer Skin) Behavior
Regardless of the damping ratio, dynamic response amplification factors (Hd) of the
secondary structure are always 0 at p = 1. As the damping ratio becomes higher, Hd
(max) of the secondary structure becomes lower. There is no optimum damping ratio
after which Hd (max) becomes higher.
7.3. Entire System Behavior
There is no single optimum damping ratio that minimizes both H (max) and Hd (max),
which is different from the situation in Tuned Mass Damper Systems.
At the optimum damping ratio for the primary structure, Hd (max) is much higher than H
(max), which is not desirable because Hd represents the displacement of the DSF outer
skin. As the damping ratio keeps increasing (passing through the optimum damping ratio
for the primary structure), Hd (max) becomes lower and closer to H (max). Even though
H (max) becomes higher after passing the optimum damping ratio for the primary
structure, this H (max) increase is not significant. Thus, the final damping ratio can be
determined by considering both H (max) and Hd (max). This process satisfies the
displacement requirements of both the primary structure and the DSF outer skin.
7.4. Limitations
1000
Hopt
1
m
800-
600-
400-
200-
0o.bl0
0.2
188
0.0b4
More importantly, when primary structure damping is considered, it turns out that the
secondary mass acts as a force amplifier instead of an energy dissipater in the region
where p is close to 1. This phenomenon occurs because the optimized f (wCd/C, where Cod
is the natural frequency of Md, and co is the natural frequency of M) value is very close
to 1 as can be seen in the optimized f vs. m plot. In this case, if the forcing frequency is
close to the primary mass frequency, it is also close to the secondary mass frequency.
Thus, the transmissibility between the force and the primary structure - through the
secondary structure - becomes bigger than 1, which is undesirable for the concept of this
1
system. (Transmissibility can be seen
from the H3plot blow.)
10 -=99
m..
_+m
0.98
'
0.97
0.960.95
0.93
. _._ _...
0.94
001b'
0.02
0.03
0.04
0.5
- .
2
L'
0.
Transmissibility H3Plot
R = H3P
R: Reaction
P: Applied Force
189
190
and TMD
Analysis
An SDOF/TMD system is analyzed here in order to investigate phase shifts between the
primary structure and TMD in various forcing frequency conditions.
Em!lishlGreek Svmbols
--p
M: Mass
K: Stiffness
C: Viscous Damping Parameter
P: Applied Loading
U: Displacement
m: Natural Frequency of Structure
n: Forcing Frequency
p: Forcing Frequency to Natural Frequency Ratio (Wm)
f: TMD Frequency to Primary Structure Frequency Ratio (md / m)
~: Damping Ratio
cjl: Mode Shape Vector
K
V\/V\
-\'\,\\,
Md
M
~~
Cd
C
U
U+Ud
It is convenient to work with the solution expressed in terms of complex quantities. The
force is expressed as
(eq.3)
" iO/
P = pe
,
where
u = ueiO/
(eq.5)
where the response amplitudes,
and
UJ,
Then the corresponding solution is given by either the real or imaginary part of
anduJ
(eq.9)
191
4-
(eq.10)
/d
Cd = 2dt)d md
(eq.1 1)
(eq.1-2)
md)du,
(-md
(eq. 1-3)
(eq.1-4)
(eq.2-2)
(-mdp02o
+i2:,ddmdpc
+w, md,
)u - mdp22u=O
(eq.2-3)
(-mp +i2,df m p+ f
(eq.2-4)
.2
)u-
ps =0.
k
Ud =
Hd
J[f 2 -P ]2 +[2rdf] 2
P2
d-=ID
192
(eq.14)
(eq.15)
(eq. 16)
(eq.17)
U+Ud
If structural damping for the primary structure is not considered, the equations reduce to
- PH
u=-
(eq.18)
PH
u- d =k
(eq.19)
dl
(eq.20)
(eq.21)
(eq.22)
For the condition when primary structure damping is not considered, the phase angle
a between the primary structure and the TMD is expressed as
tana=
2~dpf
2
f -p
(eq.23)
193
arctanK
f2 p
-i(
7
)1 II\Qj
^ 1 -- iIililll 2
. QQ70n
^1n%_
1
k. -- ,/ -, , -- , - , ,- - ",2
(
,z.
p2f
_p 2
Considering
m(3- 0-.55
dI,, _8(1
+m-)(1
- 0.5m)
5ds 1,1
(eq.24)
I-0.5m
cn be plotted for
(eq.25)
= 0.005
0.9
0.96
1.05
1.1
1.15
rho
(Graph.
194
1)
8.2. Phase Angle between Primary Structure and TMD (MotionLab Simulation)
The phase angle between the primary structure and TMD is investigated using
MotionLab simulations. For the condition when primary structure damping is not
considered and m=0.005, the system is studied 40 times using 0.05 second intervals
between forcing frequencies around the resonance zone. The results are tabulated below.
No
Tn(sec.)
Tf(sec.)
rho
alpha(degree)
1
2
3
4
6.28
6.28
6.28
6.28
7.33
7.28
7.23
7.18
0.857
0.863
0.869
0.875
10
10
10
10
11
6.28
7.13
0.881
6.28
7.08
0.887
13
7
8
9
6.28
6.28
6.28
7.03
6.98
6.93
0.893
0.900
0.906
15
17
19
10
11
12
6.28
6.28
6.28
6.88
6.83
6.78
0.913
0.919
0.926
21
24
27
13
6.28
6.73
0.933
30
14
15
16
17
18
19
6.28
6.28
6.28
6.28
6.28
6.28
6.68
6.63
6.58
6.53
6.48
6.43
0.940
0.947
0.954
0.962
0.969
0.977
34
39
45
52
60
70
20
21
22
23
24
25
26
27
6.28
6.28
6.28
6.28
6.28
6.28
6.28
6.28
6.38
6.33
6.28
6.23
6.18
6.13
6.08
6.03
0.984
0.992
1.000
1.008
1.016
1.024
1.033
1.041
80
90
100
110
120
128
135
141
28
6.28
5.98
1.050
146
29
30
31
32
33
34
35
36
37
38
39
6.28
6.28
6.28
6.28
6.28
6.28
6.28
6.28
6.28
6.28
6.28
5.93
5.88
5.83
5.78
5.73
5.68
5.63
5.58
5.53
5.48
5.43
1.059
1.068
1.077
1.087
1.096
1.106
1.115
1.125
1.136
1.146
1.157
150
153
156
158
160
162
164
166
168
170
170
40
6.28
5.38
1.167
170
Notes
195
phase angle
180
160
140
120
100
80
60
40 20
0
I
I~
f"L()
L()
t---
co
co
('f)
('f)
co
a a
('f)
('f)
t---
a
a
f"-
L()
m a
(J)
a
~
a
L()
a
~
f"-
to
to
f"-
~
~
~
~
~
~
t---
('f)
phase angle
to
1_
(Graph. 2)
Graph 2 plots 40 MotionLab simulation results regarding the phase angle between the
primary structure and TMD. Actual MotionLab simulation results are shown in the
Appendix 8.2.1.
80
60
40
20
0
-20
-40
-60
-80
196
0.9
0.95
1.05
1.1
1.15
I
..................
~,
-1-
: I: I~
...
_~
_-:
.J
J,
tr.
~i ,~
.;run:
I
...
"
.'
II,
;! ~
\L---:\1 n:"\.;r
I
II,
,_"'_,_
I..
~i
,II
! ...
.&.I
,Ii
"
....I
Tf
1
I
::
..
, .
"
<-}
...
'J
350
Tf = 12 (almost in phase)
197
_,.
__
0r'
..
0._ __ ,,
-0"
_., __ __ 0. _0 '_0.'_
15
1.5
.0.5
-1
-1
DJ
3J5
320
310
315
time in seconds
325
---
---
----
--.-
,,-,
.~u_ -- .H.
0 --. - --.
~i'~
i051muuT'u~u
"-1---.
--1'-----------1'-----------1------------1-------
:'~i
--
1. -~---. H.
c:
15
...
.
,
,
.. - --
- --
f .... -- - -- _O"t--
!:
~,:"
~,:
,.I\,
r./, :.:
:,;f
?J:.::\{:::
u.
310
315
time in seconds
DJ
320
325
.,,
,
,
,
.. 1_._._
..
._ .. _
1.5 - - ~---- - -- - --
.. _
r----
i~:w~r:~
310
315
time in seconds
DJ
320
325
--
-----
-- ,.. ---
I0: .~
--
--
----
--
---
r\.....Ii .I.T.,./.
~,:' !'I,.'j..,,'-..A',\J.'\V
-1
DJ
310
315
time in seconds
320
325
3D
310
315
time in seconds
320
325
__~__.P.
,
.
I
. __
and damper
.1.
displacements
__ -'_
..
__
.J. _
l'tlr.L .. IL'.r':!;".]
~II..:,"ul'U'>.UFUT:hT
.2 .
~oo
.:00 00.: . ~
: . ; :
~! ,
~.;
Time history.ofthe,nodal.and
damper displacements
16
;,
t4,
12
to
E
ii
14
12
to
.2
14
12
to
-2
200
\}
1.5 ..
1 --
0.5 --
-1
-1.5
-2
1.5
c:
0.5
~
'6
-0.5
-2
20
U+Ud
(\
16
16
14
12
10
/
6
6
System
4
0.9
O.
1.05
1.1
H olot . U= (P/K) H
Mass Ratio (Md/M) = 0.005
Structure Damping Ratio ~ = 0
TMD Damping Ratio ~d = 0.043
f= 0.995
rho
200
150
100
50
Hd 'Plot. Ud = (P/K) Hd
Mass Ratio (Md/M) = 0.005
Structure Damping Ratio ~ = 0
TMD Damping Ratio ~d = 0.043
f= 0.995
203
Modal displacement
profile.
Illal part.
structure
..
1
"t>
1
-0.2
0.2
0.4
0.6
ampl~ude
The two MotionLab result graphs above show two mode shapes, first and second modal
period and the modal damping ratio of the SDOF /TMD system. In the mode shape graph,
node number 1 represents the primary structure and node number 2 represents TMD. As
can be seen from the graph, the primary structure and TMD are in phase in the first mode,
while they are out of phase in the second mode.
Considering sinusoidal periodic loading applied to the first node of the system, when
forcing frequency is 6.48 seconds (first mode period of the structure), the first mode
response governs the overall response of the system.
On the contrary, when forcing
frequency is 6.13 seconds (second mode period of the structure), the second mode
response governs the overall response of the system.
Considering the first mode shape, it can be speculated that the primary structure and
TMD motions become in phase when the first mode response governs. Likewise, it can
also be speculated that the primary structure and TMD motions become out of phase by
around 180 degrees when the second mode response governs. However, as was observed
from Graphs 1 and 2, phase angles are close to 90 degrees in both cases.
As can be seen from the Appendix 8.2.1, zero degrees in phase and 180 degrees out of
phase occur in extreme forcing periods. A zero degree phase angle occurs when the
p (Forcing Frequency to Natural Frequency Ratio (Q/m)) value is close to zero, while the
180 degree phase angle occurs when the p value is close to around 2 (infinity more
exactly speaking).
204
TMD Systems
2DOF System with Multiple Distributed TMDs at both the first and the second Nodes
- Tuned bv Iteration
d2Md2
/\/\
C d2
t:
U2~
Ud2
u, C,
;: P1
Ml = M2 = 40,000,000 kg
Mdl = Md2 = 250,000 kg
Cl = C2 = 2,000,000 N-s/m (~= 0.01)
Cdl = 35,000 (~d = 0.070)
Cd2 = 35,000 (~d = 0.071)
Kl = 120,000,000 N/m
K2 = 80,000,000 N/m
Kdl = 249,000 N/m (fl = 0.998)
Kd2 = 245,000 N/m (f2 = 0.990)
PI = 0
P2= 4,500,000 sin(wt) N
Since optimal parameters for vertically distributed TMDs cannot be analytically obtained,
the parameters used above are based on trial and error. Three different forcing
frequencies were used for the simulation: forcing frequency = first mode frequency;
forcing frequency = second mode frequency; and forcing frequency = third mode
frequency. The system was tuned until peak displacements of the primary structure
became the same in all three forcing frequency conditions. Because tuning was done only
by trial and error, it is hard to say that what is shown here is the absolutely optimized
condition for the system. Better tuning parameters, which reduce even more the dynamic
response of the primary structure, may exist.
As can be seen from the plots of the time history of the nodal and damper displacements,
phase angles between the primary structure nodes and TMDs are around 90 degrees in all
three forcing conditions.
205
200
205
Forcing Frequency
210
215
lime in seconds
220
- damper,1
,<
c.
-1
-3
200
Forcing Frequency
Forcing frequency
206
the TMD changes as the forcing period changes. When the p (Forcing Frequency to
Natural Frequency Ratio (Q2/a)) value is close to zero the phase angle becomes close to
zero, while when the p value is close to 2 (infinity, more exactly speaking), 180 degrees
out of phase occurs. Around resonance condition (p value is close to 1, around peak
points region in both H and Hd plots), which is the primary concern of the TMD design,
the phase angle between the primary structure and TMD is around 90 degrees.
In a 2DOF/2TMD system, phase angles between the primary structure nodes and TMDs
around resonance condition are still around 90 degrees. However, since optimal
parameters for vertically distributed TMDs cannot be analytically obtained, they are
based on trial and error. For this reason, what is shown here might not be the optimized
condition for the system. Better tuning parameters, which reduce even more the dynamic
response of the primary structure, may exist.
207
208
Appendix 9. 2 DOF + 2 TMD H, Hdl & Hd2 for the Approximate Solution
f1.fl
ml p +f12 ml) Udl -x ml p2 U=0
+(-p
m2) Ud2
P
+ 2 iC p+ 1) U=-
English/Greek Symbols
m: Mass Ratio
2flf2 p2 _2 2 p2 +ft2 2 )
(1/2)
p3
+ (-.2 lfJ
- 2
22
p3 +
2 2/f`22 p + 2 1 fJ2
//, ((_p2jf2
p) )
2 2 m2
4fl 2
p4 C
2 2f2
2
+ 4 4 C 2f + 4 1 7 p4 f2 m2
+4
`7 p4 ,22`2+ p4) 1+ (2 p C2
C+ 2 pf17 1
2
3
2
3
3
5
2
-2
l fl p + 2 p u fl 111l 1 - 2 p a2fl ml 1f22 _ 2 I( f1 ml 52f2
5
+ 2 p 42 2 - 2 51 7 p3 .22 _ 8 51 7 p3 C 2f2 - 2 f 2 p3 f2 m2 42 + 2 fl 2 C p j2 2
+2 2/2f22
- 21
p322 m2 +2
2 ()1/
)2
ff22 p - 242J2p) )
209
Hdl := (
p4 _f
p2f22
+ p4j2 2
2 ml
m2 - p6 +j 222 + p4j2 2 _ p2J22 + 4 p4 a2fl ml 4152J2 + p4 2 Jf7
- P2 fl2mif22 + 4 1fl p4 C'-41 fl p2Cf22 - 4 lfl p 2 f2
-.f72 2./22m24fl2 C 52.2
22p2+ 44 C 2f/ + 4 ,lf7 p4f2 m2 ,2
2
+451fi p452f2+ p4) +(2pSC+ 2 p5fl 1 -241f
p3+2 p5c2j 7 ml 41
- 2 p3 U2 fl ml1 12 - 2 p3 c 22 M2 52 2ml
+ 2 p 5 42f2 - 2 1fl p3 f2 2
- 8 41f/ p3 C 42.f2 - 2 f 2 p3 f2 m2 42 + 2 f12 C p f2 2 + 2fl 2 42.f2p
- 2j12 p3 52f2 - 2
2 C P3 + 2 p 5 f2 m2 2 - 2 p3 C2 2 - 2 ,1fl p 3 .f22 m2
2 (1/2)
p3
+ 2 1.f7 p f2 2
2f2)
2 p2) + 4 12j2 p6
Hd2:= (p 4-_f7
_
2 2.f22+p 412 2 m2
p6 + j2j222 + p4 j2 2 _ p 2 f2 2 + 4 p4 (2J ml 1 2J2 + p4 02Jf2 ml
2 C2 ml2 2 + 4 1fl
C- 4 1fl p 2 C2 2 _4 1fl p2 2f2
-_Jf2p2 f22
4f2
((-
C p2 2.f2 + 4 p4 C
2 +f2p
2f2 + 4
1 fZ p4f2 m2 2
+ 2 ,1f7 pf2 2
210
p 3 2f/2) )
5
4
3
0.5
1.5
rho
10
8
6
4
2
50
(\
40
~
30
2D
10
)
0
0.5
211
Optimal parameters for minimum HA are slightly different from those for minimum H.
However, this difference is not significant.
12-
HAI plot,
IIl
11-=0.5, 12=1.0
10-
21 = 1.0, 22 = -0.5
8-
0)1 =1
0)2= 2.5
Mass Ratio M1 = 0.015
64-
/
..
"-\
21
1.5
rho
2.5
fl = 0.985, f2 = 980
TMD Damping Ratio 4dl = 0.065, d2 = 0.095
Ml
11 = 0.5, 12 = 1.0
)21 = 1.0, 22 =-0.5
B'
6'
1ol = 1
42U
0.5
. =-r-.-
-2-4-
212
....
0:5
.....
1.'5
rho
I'
.,..,f
)2 = 2.5
Mass Ratio M = 0.015
Mass Ratio M2 = 0.015
Struc. Damping Ratio 1 = 0.005, 2 = 0.0132
fl = 0.985, f2 = 980
TMD Damping Ratio dl = 0.065, d2= 0.095
0.5
1.5
rho
30
25
20
15
10
5
0.5
1.5
rho
2.5
213
Optimal parameters for minimum HA are slightly different from those for minimum H.
However, this difference is not significant.
- I
. . I
HA1 plot A1= (/M) HA!
S.
5.
4-
1ol
=
o2 = 2.5
Mass Ratio M1 = 0.003
Mass Ratio M2 = 0.030
3.
2,
2 = 0.0132
fl =2.51,f2 =0.985
U
0.5
.,
I
1.5
rho
2.5
'is'
'.i
rho
214
d2 =
0.11
(ol=
(o2= 2.5
Xl;
'5rho
:I
'2.5 /
1
1
0.98
1.02
1.04
rho
400-
300
200
0o
1.99
JI
0.992
1.
0.96
0.994
-.
' 0.998
rho
I ~}I
I L-l
! ~
l~-I
M(O.33%)
- M(O.33%)
M (0.33%)
MotionLab Tn
in
6~0691
O.:~:4 I O:'~~8
.Optimized Condition
[~}I
l~-I
I ~-I
M(O.33%)
M(o.33%)
M(O.33%)
-~
Cd3
Cd2
Cd1
=
=
=
13,593N-s/m
11,144N-s/m
8,685N-s/m
~ = 0.005
216
0.055)
= 0.045)
= 0.035)
(~d3 =
(~d2
(~dl
1.034
6.56
11.2. Sensitivity of the System to the Change of the Primary Structure Damping
a l = 1/3, oc2= 2/3
50
40
30
20
10
U..u
rho
50-
40-
30-
20-
10i
. .2
-
0 1.99'
_
.
0 .992
LI-,.
I
-- )_, 11------I.
.
.
0.994
0.996
.
.
0.998
..
'i
rho
217
= 1/3, c2 = 2/3
1 = 0.001, 2 = 0.001,
3 = 0.001, ,= 0.01
40
30
20
10
1.04
rho
2.5//
(
2//
1.5-
I"*
1`I,
'=
0.99
'g
0.992
0.996
0.994
rho
218
0.998
+ cu + ku
of TMDs
IV
c..
--j
Figure (a)
Figure (b)
+ Cd(n-l)Ud(n-l
+ (kdnUdn + cdnudn)
mdludl
+ cd(n-I)Ud(n-l) + kd(n-l)Ud(n-l)
+ cdnudn + kdnUdn = -mdnu
md(n-l)Ud(n-l)
mdnudn
= -md(n_l)U
mu
+ CU + ku
P + kdud + CdUd
mdud
Thus, the model in Figure (a) composed of multiple TMDs can be simplified into the
model in Figure (b) composed of a single TMD, and the following simple relationships
exist between the two models.
=
= md(n-l) = mdn = md
kdl
= kd(n-l) = kdn
Cdl
= cd(n-l)
Udl
= ud(n-l) = Udn = Ud
mdl
= kd / n
= Cdn = Cd / n
219
220
50
40
.n
/\
30
H plot. U= (P/K) H
20
~=O
TMD Damping Ratio
~d = 0.01, 0.03, 0.05, 0.07,
0.09, 0.2
f= 0.99
10
0.8
0.9
1
rho
1.1
500
400
300
Hd plot. Ud
200
(P/K) Hd
~=O
TMD Damping Ratio
~d = 0.01, 0.03, 0.05, 0.07,
0.09, 0.2
f= 0.99
100
0.8
0.9
1.2
rho
221
25
20
15
'
/ \'.
i
!
H plot. U= (P/K) H
Mass Ratio = 0.01
Structure Damping Ratio
~ = 0.01
TMD Damping Ratio
10
~d
0.9
0.8
1.1
1.2
rho
250
200
Hd plot. Ud = (P/K) Hd
150
100
50
0.8
0.9
1
rho
222
1.1
1.2
16
14
12
10
H plot. U= (P/K) H
8
6
4
2
0.8
0.9
1.1
1.2
rho
180
160
140
120
Hd plot. Ud
100
(P/K) Hd
80
60
40
20
0.8
0.9
1.1
rho
223
7
6
H plot. U= (P/K) H
5
4
3
2
0.8
0.9
1
rho
1.1
1.2
100
80
60
Hd plot. Ud
40
20
0.8
0.9
1
rho
224
1.1
(P/K) Hd
Part I References
Abalos, Inaki and Juan Herreros. Tower and Office. From Modernist Theory to
Contemporary Practice. Cambridge: MIT Press, 2003.
Basalla, George. The Evolution of Technology. New York: Cambridge University Press,
1988.
Banham, Reyner. Theory and Design in the First Machine Age. London: Architectural
Press, 1960.
19 th Century.
Press, 1961.
Condit, Carl. American Building Art: The 20th Century. New York: Oxford University
Press, 1961.
Condit, Carl. American Building: Materials and Techniques from the First Colonial
Settlements to the Present. Chicago: University of Chicago Press, 1968.
Connor, Jerome. Introduction to Structural Motion Control. New York: Prentice Hall,
2003.
Fernandez, John. MIT 4.463: Advanced Structures and Building Envelopes, Lecture Note.
MIT, 2004.
Gidion, Sigfried. Space, Time and Architecture: The Growth of a New Tradition.
Edition. Cambridge: Harvard University Press, 1967.
th
225
Landau, Sarah and Carl Condit. Rise of the New York Skyscraper, 1865-1913. New
Haven: Yale University Press, 1996.
Letherbarrow, David. Uncommon Ground Cambridge: MIT Press, 2000.
Messler, Norbert. The Art Deco Skyscraper in New York. New York: P. Lang, 1986.
Loughran, Patrick. Falling Glass. Problems and Solutions in Contemporary Architecture.
Basel: Birkhauser, 2003.
Pevsner, Nikolaus. The Source of Modern Architecture and Design. New York: F. A.
Praeger, 1968.
Robinson, Cervin. Skyscraper Style. New York: Oxford University Press, 1975.
Rowe, Colin and Robert Slutzky. Transparency. Basel: Birkhauser Verlag, 1997.
Russell, James. Architectural Style and Management Ideals. AV Monographs 103, 2003.
Schueller, Wolfgang. The Vertical Building Structure. New York: Van Nostrand
Reinhold, 1990.
Smith, Bryan and Alex Coull. Tall Building Structures: Analysis and Design. New York:
Wiley, 1991.
226
Willis, Carol. Form Follows Finance: Skyscrapers and Skylines in New York and
Chicago. New York: Princeton Architectural Press, 1995.
Part II References
AISC. Manual of Steel construction. Load & Resistance Factor Design. American
Institute of Steel Construction Inc., 1998.
ASCE. Minimum Design Loads for Buildings and Other Structures. American Society of
Civil Engineers, 1996.
Abe, M. & T. Igusa. "Tuned Mass Dampers for Structures with Closely Spaced Natural
Frequencies." Earthquake Engineering and Structural Dynamics. vol.24, no.2. Feb. 1995.
247-61.
Abe, M. & Y. Fugino. "Dynamic Characterization of Multiple Tuned Mass Dampers and
Some Design Formulas." Earthquake Engineering and Structural Dynamics. vol.23, no.8.
Aug. 1994. 813-35.
Ankireddi, S. & H. Yang. "Simple ATMD Control Methodology for Tall Buildings
subject to Wind Loads." Journal of Structural Engineering. vol.122, no.1. Jan. 1996. 8391.
The Chicago Committee on High Rise Buildings. Exterior Claddings on High Rise
Buildings Past Experience, Present Directions, and Future Trends. The Chicago
Committee on High Rise Buildings, 1990.
Clough, R. & J. Penzien. Dynamics of Structures. New York: McGraw-Hill, 1975.
Compagno, Andrea. Intelligent Glass Facades. Basel: Birkhauser, 1999.
Council on Tall Buildings and Urban Habitat. Architecture of Tall Buildings. New York:
McGraw-Hill, 1995.
Council on Tall Buildings and Urban Habitat. Tall Building Criteria and Loading.
Council on Tall Buildings and Urban Habitat, 1978.
227
Council on Tall Buildings and Urban Habitat. Advances in Tall Buildings. Council on
Tall Buildings and Urban Habitat, 1980.
Council on Tall Buildings and Urban Habitat. Habitat and The High-Rise: Tradition and
Innovation. Dutch Council on Tall Buildings, 1995.
Council on Tall Buildings and Urban Habitat. Cladding. New York: McGraw-Hill, 1992.
Council on Tall Buildings and Urban Habitat. Structural Systems for Tall Buildings.
New York: McGraw-Hill, 1995.
Council on Tall Buildings and Urban Habitat. Tall Buildings: 2000 and Beyond Council
on Tall Buildings and Urban Habitat, 1990.
Council on Tall Buildings and Urban Habitat. Tall Building Systems and Concepts.
American Society of Civil Engineers, 1980.
Den Hartog, J. Mechanical Vibrations. 4th Edition. New York: McGraw-Hill, 1956.
Kim, Michael. Arch. 444: Building Systems and Design Integration, Class Note.
University of Illinois at Urbana-Champaign. 1998
Kim, Michael. Arch. 445: Design and Constructability, Class Note. University of Illinois
at Urbana-Champaign.
1997
Schueller, Wolfgang. The Vertical Building Structure. Van Nostrand Reinhold, 1990.
Taranath, B. Steel, Concrete, & Composite Design of Tall Buildings. New York:
McGraw-Hill, 1998.
228
Xu, K. and T. Igusa. "Vibration Control using Multiple Tuned Mass Dampers." Journal
of Sound and Vibration. vol. 175. 1994. 491-503.
Yamaguchi, H. and N. Harnpornchai. "Fundamental Characteristics of Multiple Tuned
Mass Dampers for Suppressing Harmonically Forced Oscillation." Earthquake
Engineering and Structural Dynamics. vol.22. 1993. 51-62.
229
230