Building

*t_
David Macaulay f4 i^
I)ompanion to the PBS® series
 ;^.5U.UU -.*?•

Why this shape and not that? Why

steel instead of concrete or stone?

Why put it here and not over there?

ii.... r.r" the kinds of questions that
David Macaulay asks himself when
he observes an architectural wonder.
These questions take him back to the
basic process of design in which all
structures begin, a process that
proceeds from the realization of a
need for the structure to the struggles

of the engineers and designers to map

out and create the final construction.

David Macaulay, as only he can,

engages readers' imaginations and
gets them thinking about structures
they see and use every day bridges, —
tunnels, skyscrapers, domes, and
dams. In Building Big he focuses on
the connections between the planning
and design problems and the solutions
that are finally reached. Whether a
structure is imposing or inspiring, he
shows us that common sense and
logic play just as important a part in
architecture as do imagination and
technology. As always, Mr. Macaulay
inspires readers of all ages to look at
their world in a new way.

rhe public television series Building Big is a coproduction of

he WGBH Science Unit, producers of the award-winning

\ova science series, and Production Group, Inc.

1000
 Building Big

C5 «'
^- '

^
http://www.archJye.Qrg/details/buildingbigpQn5abar/)^
^3V
/\r/)
Building Big; ]
David Macaulay /y-//

^^\
X
HQIJgllTON MIFFLIN COMPANY BOSTON 2000,

'Walter Lorraine (^ Books

r^- Jl^
 For their willingness to review and comment upon various
parts of this book, I would like to thank the following: Chuck
Seim, Gary S. Barry Cooke, William C. Allen,
Brierley, J.

William L. MacDonald, Professor Mark Cruvellier and

Executive Producer of the series, Larry Klein, who, unhin-
dered by expertise in any of the subjects, read them all. At
Houghton Mifflin, my gratitude as always to my editor, Walter
Lorraine and to Donna McCarthy, Margaret Melvin, Liz
Duvall and Kathy Black, none of whom ever gave up even as
the deadlines came and went. Finally, thank you Ruthie for
your unfailing support and for putting up with the absences,
the frustration and the endless same-old-conversations with
humor and unshakable confidence. This book is for you, love.

Walter Lorraine (^ Books

Text and illustrations copyright © 2000 by David Macaulay

All rights reserved. For information about permission

to reproduce selections from this book, write to

Permissions, Houghton Mifflin Company 215 Park
Avenue South, New York, New York 10003.

Library of Congress Cataloging-in-Publication Data

Macaulay, David.
Building Big / by David Macaulay
p. cm.
ISBN 0-395-96331-1
1. Architecture-Juvenile literature. [1. Architecture.] L Title

NA2555.M24 2000
720—dc21
00-028116
OH BR
J
Printed in the United States of America NA2555
RMT 10 987654321 .M24
2000
 Contents
Preface 7

Bridges 9
/ Ponte Fabricio 11
Iron Bridge 15
Britannia Bridge 16
Garabit Viaduct 20
i Firth of Forth 24
Golden Gate Bridge 30
Ponte de Normandie 52

Tunnels 57
Two Ancient Tunnels 58
Hoosac Tunnel 60
Thames Tunnel 69
Holland Tunnel 76
The Channel Tunnel 78
The Big Dig 87

Dams 93
ItaDam 94
Hoover Dam 111

W P Aswan High Dam

Itaipu Dam
118
122

Domes 129
Pantheon 131
Hagia Sophia 136
Sehzade Mosque 140
St. Peter's Basilica 141
Les Invalides 144
St. Paul's Cathedral 144
United States Capitol 145
Astrodome 154

Skyscrapers 161
Reliance Building 162
Woolworth Building 166
Chrysler Building 167
Empire State Building 168
John Hancock Center 174
World Trade Center 175
Sears Tower 176
Citigroup Center 179
Petronas Towers 180
Commerzbank Frankfurt 186

Glossary 192
 i ^^^^iVXiXK^iSS

\t
1 ' 1 ^
l5<
^^

•I 1 1 A
iiaiatl

W""ll lllllll'^l
 —

Preface
Building Big began as five films about the creation of
bridges, tunnels, skyscrapers, domes, and dams. Over
a period of about two years, various producers, film
crews, and I checked in and out of hotels on four con-

A tinents and talked to a lot of people who design, build,

or study these structures. While the filmmakers were
concentrating at least as
stories
much on
—the ambition, the heartbreak, the triumph
the human

as on the technical,
X found myself increasingly
intrigued by the nuts and bolts.
I

It's just the kind of

person I am. Why this shape and not that? Why steel
instead of concrete or stone? Why put it here and not
over there? Asking these kinds of questions took me
back to the basic design process, which itself begins
with questions as engineers and designers struggle to
identify and prioritize the problems that must be
solved. And it was this particular aspect of building
big that finally convinced me there was a role for a
so-called companion book.

Knowing that the films would present the big

picture — including the larger historical, social, and
environmental issues associated with the building of
—
big things I was free to operate on a much smaller
scale. Using some but not all of the examples chosen
for the films, this book focuses entirely on the
connections between the main planning and design
problems that had to be solved and the solutions that
were eventually built. There is something reassuring
about the fact that whether structures inspire or
simply intimidate with their scale, each is generally
the result of a logical and therefore accessible
sequence of events. Once we recognize that the
elements of common sense and logic play at least as
important a role in this process as imagination and
technical know-how, even the biggest things we build
can be brought down to size.
 r

T—

^
 Bridges
All the structures in this book willingly reveal
important things about why and how they were built
if we know what to look for. And of all big pieces of
engineering, bridges are probably the most forth-
coming. They are in a sense three-dimensional
diagrams of the work they do, and this makes them
ideal subjects with which to begin.
In large modern bridges, where economy is of the
essence, there are very few extra elements to obscure
what's going on. What gives these bridges any
uniqueness they may have is not some applied
decoration but rather their fundamental design and
its relationship to the site. Even in the often conser-

vatively built and frequently embellished bridges of

old, the way they work can be readily appreciated.

While the specific requirements of every bridge are

different, thereby creating an enormous number of
variations, there are actually only five basic bridge
types — postand beam, arched, cantilevered,
suspension, and the newest kid on the block, cable
stayed. The bridges included here were not necessar-
ily the largest of their day, but like all bridges,
whether dramatic or humble, they all evolved from a
similar process. This included defining the problem,
establishing the goals, and testing the limits. In the
end, each completed structure either reflects the
technological understanding of its day or represents
a leap in that understanding.
The other thing these bridges have in common is

that they were span water a challenge

all built to —
that seems to have an enduring appeal to those of us
with neither wings nor gills. Over the centuries, this
problem has been met with ingenuity, common sense,
and courage and has resulted in some of the most
magnificent pieces of engineering on the surface of
the planet.

\
iW JK

,^rm"-
 PONTE FABRICIO
Rome, 62 B.C.: The task of the road commis-
Italy, stone arch: absolutely permanent, absolutely fire-
sioner and his engineers was to build a bridge linking proof,and absolutely unclimbable.
one bank of the Tiber River with the island in the Whether any of these plans was seriously consid-
middle, a distance of approximately 200 feet. Because ered is unlikely, given the level of experience the
the island housed medical facilities, it was crucial that Roman engineers had. The final design was a stone
people could cross the bridge easily, so the bridge structure with two arches. It offered permanence (in
couldn't be too steep. There was a constant stream fact, it still stands today), created only one impedi-
of boat traffic from the port of Ostia, so the bridge ment on the river, and was high enough to allow ships
couldn't be too low either to pass underneath yet low enough to provide an easy
The kinds of construction materials that may have foot crossing. The three small arches were added to
been considered include an all-wood post-and-beam reduce pressure against the bridge during flooding.
structure (inexpensive but not fireproof), timber The commissioner apparently believed that his
beams on a single stone pier (not entirely fireproof engineers had come upon a worthy solution: he had
but less of an obstacle to river traffic), and a single his name carved in four places on the bridge.

'^
M

c\ >-n Ti» r'( n-m -r. -m— t-

' i. ,-^- ,
'
_ li .

11
 Roman engineers understood that a structure is

only as permanent as the foundations upon which it

stands. The building of foundations at any site is

tedious, but to build them in the middle of a river can
be downright treacherous. The builders probably
waited until the summer, when the river was low, to
begin the Ponte Fabricio. They may have diverted the
water using some kind of temporary barrier. Or they
might have built a watertight wooden wall, called a
coffer dam, around the site of the central pier. Once
pumped dry, the enclosed space could be excavated
down to the riverbed.

tree
trunks

clay
infill

Coffer Dam

Next to the foundations inimportance is the pair of

arches. These would have been built over a tempo-
rary wooden form called a centering. Once the last
piece of stone, called the keystone, had been set in
place, the arch proper was finished. But if the center-
ing was to be removed at this point, the uppermost
stones would want to fall (it's one of the things stones
do best), and this in turn would push the sides of the
arch outward. To prevent this from happening, more
material was built up against the sides of the arch
before the centering was removed. This not only
prevented movement between the stones but also
compressed them together. The more the wedge-
shaped stones are compressed, the stronger the arch
will become. Thus, it is capable of channeling all the
weight from above to the foundations below.
The fact that the bridge is more than 2,000 years
old and still in use is a testament to the import-
ance of combining the right shape with the right
materials.

13
 Iron Bridge
Coalbrookdale, England, 1775: In the second half of Unlike its masonry predecessor, Ponte Fabricio,
the eighteenth century, the Severn River, which Iron Bridge was assembled like an oversize Erector*
runs through Coalbrookdale, England, was a major set of about 800 pieces. For each piece a full-size
obstacle to getting raw materials from one side of this wooden pattern was pressed into a bed of sand and
industrialized valley to the other. There were very then carefully removed. The resulting cavity was
few bridges because of the heavy traffic on the river, filled with molten iron. Using traditional carpentry

so materials were transported on ferries. However, joints, such as dovetails, mortises, and tenons, the
the ferries could not keep up with the increasing builders would have put the finished pieces together
productivity of the area. from a fairly lightweight scaffold instead of a heavy
A new bridge was needed, and it would have to be timber centering, which would have brought river
arched, so as not to block river traffic. Potential traffic to a standstill. While the arch was clearly the
builders were asked to think in terms of wood, stone, best design to span the river, it also turned out to be
or brick. But the iron master Abraham Darby III saw the best shape for the material. An arch is a com-
the challenge as an opportunity to promote his own pression structure, and cast iron, like stone, is at its
skills and foundry. Stone was hard and heavy,
to cut happiest when it is compressed. Whether Darby knew
which made it expensive to work with and to trans- this or was simply lucky is anyone's guess. The bridge
port. Iron was more efficient to build with because it still stands today because it was designed with the

could be cast into the exact shapes required and much right combination of shape and material. Also, the
closer to the building site. Darby built the first all- openness of the structure allows floodwater to pass
metal, prefabricated bridge in the world. easily underneath it.
 Britannia Bridge
Bangor, North Wales, 1838: The task for railway was a minimum height requirement of 100 feet
engineer Robert Stevenson was to build a bridge both between bridge and water.
strong enough and rigid enough to carry trains back After considering various locations, Stevenson
and forth between the Welsh mainland and the island chose, not surprisingly, a site that offered the well-
of Anglesey. There was another obstacle besides the placed advantage of a small island near the middle of
900 feet of water that separates the locations. Because the strait. Nobody said he couldn't use what was
the bridge was to be built over a working shipping already there. And he could think in
at least now,
channel, the British navy would have to approve terms of two spans rather than one long one. Since
Stevenson's design. Arches and piers were forbidden arches were not an option, he designed a bridge based
for fear they might constrict the waterway, and there on the post-and-beam structure.
 compression

To understand how a beam works, imagine a plank

placed over two supports. If you stand on the middle
of the plank, it will bend. The top surface of the plank
gi'ows slightly shorter because it is being compressed;
the bottom surface, on the other hand, is being
stretched because it is in tension. If the plank cannot
withstand one or both of these forces, it will break.
While the piers of Stevenson's new bridge were
under construction, another of his post-and-beam
bridges on the same train line collapsed, sending five
people to their deaths in the River Dee. The bridge's
cast-iron girderswere too shallow to resist the
bending. They began to twist sideways and quickly
buckled. Although this was a very unusual failure,
Stevenson and his fellow engineers learned that in
addition to bending, beams must also be able to with-
stand twisting.

•:-\^%^

The Dee Bridge, May 24, 1847

—
I
compression

To minimize bending over a 460-foot span, massive

beams would have to be used. Wooden beams were out
of the question for a variety of reasons, and iron
beams would have been too heavy and cumbersome.
But now imagine our plank turned on its edge. Again
the top surface of the plank will be in compression
and the bottom in tension, but there is far less
bending, because the bending stress is spread out over
W r tension
Zf
a deeper area. Stevenson, working with fellow engi-
neer William Fairburn, solved the problem by making
the beam large enough for trains to go through rather
than over.

17
 —
I ;
II ii II t-'^-H-T-'

'

—
n
f
I

ii
II

' II
II

11
I
II
J_
II

'

rT" ii y -r I I
•

c /
 The finished bridge was designed as two parallel spanned the water were fabricated on the banks
tubes, each 30 feet high and 15 feet wide. They were and floated into position at high tide. Once fitted into
to be made of wrought iron and supported on high the vertical slots of their respective piers, they were
stone piers. While the sides were primarily flat plates, slowly and carefully lifted by powerful jacks to the
the top and bottom were assembled as smaller paral- required height.
lel tubes. All the pieces were riveted together. Because The Britannia Bridge is the only bridge described in
wrought iron, unlike cast iron, works about as well in this chapter that no longer standing. The bridge
is

tension as in compression, more material was used was constructed of fireproof materials, but a fire in
at the top of the beam than at the bottom to help 1970 distorted the metal. The bridge was no longer
prevent the kind of buckling failure that brought straight enough to carry trains and therefore was
down the Dee Bridge. The four main tubes that replaced by one with arches!
 Garabit Viaduct
Saint Flour, France, 1879: Another country, another area which, in addition to rugged terrain and deep
train bridge over water —this time to carry a freight gorges, included very strong winds.
line through the Massif Central. The body of water to Rather than simply using sheer mass to defeat
be crossed was a mere stream compared to the Menai the wind, Eiffel designed instead to outsmart it. He
Strait, in Wales; however, this stream was 400 feet created open airy structures through which the wind
below the intended line of the track. The engineer was could pass more easily and which required less mate-
Gustave Eiffel, and Garabit was to be his last bridge. rial to build —another important consideration in
He had designed hundreds of them, including railway these remote locations
bridges in this very region. He understood both the The structure of the Garabit viaduct is based on
demands of trains and the natural conditions of the the truss —basically a collection of interconnected
triangles the sides of which carry the tension and
compression compression forces.

either in tension or
compression depending
on location of train
compression tension
PRATT TRUSS FINK TRUSS

Proposed Rail Line of Garabit Viaduct

Truss bridges were commonly used in nineteenth- Two of the trusses below are named for their inven-
century North America to carry trains west. Because tors, the other two for their shape. Most trusses are
wood was plentiful and could be worked with basic built with a slight upward curvature called camber.
carpentry skills, these bridges were quick to build. When a live load, such as a train, pushes down on a
However, they were also quick to burn. As the bridge, the truss straightens out but never sags.
workings of the truss became better understood,
compression
various timber pieces were replaced by stronger iron.
Today, steel trusses are among the most common
bridges in North America. They are very strong and
can span long distances.
compression

CAMELBACK TRUSS H~T7irT

^£^^r.^^^3
as^fl^^
'.^^^^^^H
^^2E2
Eiffel's solution is basically a post-and-beam design
some 1,850 feet long. As the gorge begins to descend,
the track is carried from stone arches to a long
straight truss supported on a row of towers. Instead
of building three high towers in the center, Eiffel
spanned the deepest part of the gorge with a 530-foot-
wide arch. Two smaller towers rest on the arch, while
itscrown serves as the third support.
To permit natural expansion and contraction, the
ends of the arch are hinged rather than rigidly fixed
to their masonry supports. To increase stiffness, the
bases of the towers and arch are wider at the bottom
than at the top. To minimize wind resistance and to
reduce weight, the pieces of the truss are built of even
smaller pieces —wrought iron strips and angles all riv-

eted together. The two halves of the arch were built

out, or cantilevered, from their hinges and held in
place by cables until they met in the center. No
centering was needed, but a temporary wooden bridge
was built to carry workers and materials more
efficiently across the floor of the gorge.
 Firth of Forth
The Tay Bridge Disaster, December 28, 1879

scrapped.The task now fell to two engineers, John

Fowler and Benjamin Baker They were not only to
design a bridge that would cross the Firth of Forth but
also to rebuild the confidence of train passengers.
Their structure had to be strong, and it also had to
look strong.
After surveying the site to record the configuration
of the riverbed, the builders established the line of the
bridge across the Firth between North and South
Queensferry, a distance of about a mile and a half.

Roughly two thirds of that distance was over water. A

post and beam bridge was ruled out early. Even if the
piers hadn't created a problem for river traffic, the
depth of the water, which reached 220 feet at a couple
of places, would have made their construction impos-
sible. A suspension bridge was also briefly considered.

But while it would have supported the roadbed well

above the water, it was not seen as reliable enough to
demands of either trains or passengers.
satisfy the
The engineering team finally settled on a design
that would employ three enormous cantilevered
South Queensferry, Scotland, 1880: On the night of sections and two smaller suspended sections. The
December 28, 1879, part of the Tay Bridge, a multi- center cantilever would stand on the small island of
spanned post-and-beam structure across the Firth of Inchgarvie, in the middle of the estuary. The southern
Tay, collapsed during a severe gale, plunging into the cantilever would be located as close to the center one
water with a loss of 70 lives. The reputation of as the depth of the water would allow. That specific
the bridge's designer was ruined, and his plans for distance would then be used between the center
another bridge on the same line were promptly cantilever and the northern cantilever.

Proposed Rail Line for Firth of Forth Bridge

South Queensferry
 A cantilever is a horizontal beam fixed at only one cantilever
end. Generally, two cantilevers face each other to
support a suspended span. Stevenson's Britannia
Bridge illustrated how by increasing the depth of a
beam it is possible to reduce the bending. Since most Ih'Nimi
of this bending happens in the center of the beam,
that is where the depth should be greatest. In the can-
tilevers of the Forth Bridge, it's as if the supports of
the beam have been moved into the center below the
heaviest and hardest working part of the beam. With
less weight at the points farthest from the supports,
there is less bending and therefore, less building

material is required. Here the cantilevered beams

extend in opposite directions from each of the three
towers to maintain a balance. If we link these three
structures together by suspending two smaller
bridges between them, we have a continuous beam.
The separate sections are actually hinged together so
that the largest forces will still be channeled directly
to the foundations.

hinge

- Inchgarvie Island '

North Queensferry
 Each tower has four tubular steel legs, and each leg Men could now work reasonably safely on the
stands on its own massive pier. Of the twelve legs riverbed. As they excavated muck and loose rock,
needed, six had to be built in water so deep that pneu- sending it up one of the shafts, the space above them
matic caissons were required. These wrought-iron was filled with concrete. This increased the weight of
cylinders were 70 feet in diameter and roughly 70 feet the caisson, which slowly sank on its wedge-shaped
high. They were brought to the site in pieces and base. Once a firm footing had been reached, the work
reassembled on the beach before being towed to their chamber too was filled with concrete. Now there was
final locations and sunk. The work chamber at the a secure base on which to build the masonry piers that
bottom of the caisson and the work platforms at the would support the steelwork. Half of the construction
top were linked by three tubular shafts two were — time for the entire bridge was spent building the
used to transport materials and one was used by foundations. Fabricius would have approved.
workers. Each was accessed through an airlock. Water
at thebottom of the work chamber was forced out as
compressed air was pumped in.

permanent caisson

concrete

work chamber
(J /

ti^\ {^s^=*^=^
u ml
Completed Firth of Forth Bridge tower cantilevered truss suspended span
 The forces of the wind as well as those created by a
crossing train are carried through the cantilevered
trusses to the three huge towers, the main tubes of
which are twelve feet in diameter. The five tubes that
meet at each corner of a tower are riveted together
and tied into the pier below. As at Garabit, the entire
structure is wider at the bottom than at the top for
greater stability.
Fowler and Baker created a structure so strong that
the wind could never blow it down. And one only

needs see how totally dwarfed a train is by the bridge

to imagine the reassuring psychological impact it

must have had on even the most skeptical passengers.

But Fowler and Baker also created a structure so
incredibly expensive that it has rarely been copied,
particularly at this scale. The only cantilever bridge
larger than the Forth is a visually inferior imitation
built in Quebec in 1917.

"^ ^:P;:i ^P^<^

29
 Golden Gate Bridge
San Francisco, California, 1930: The problem was too place to build the bridge, it had evolved into a pure
many cars and not enough ferries to get them from suspension design.
the city to the spectacular countryside of Marin The main parts of a suspension bridge, other than
County and Northern California. People were waiting the roadway, are the towers, the cables, and the cable
—
in line for hours sometimes even days. The solution anchorages. The roadway itself actually hangs from
was a bridge. A number of locations were scouted, but the cables. If the cables stopped at the tops of the
the Golden Gate site was chosen because it would towers, their own weight in addition to the weight of
require a shorter span and a less extensive network of the roadway and that of the traffic would bend the
approach roads than any of the others. Still, the span tops of the towers toward each other. To prevent this
would have to be longer than anything built before it. from happening, the cables pass over the tops of the
An engineer by the name of Joseph Strauss was eager towers and are connected to concrete anchorages cast
to take on the challenge, although his first proposal into solid rock. The action of the cables pulling down
was an ungainly combination of cantilever and on both sides of a tower creates a strong vertical force
suspension bridge. By the time the funding was in that must be carried to the foundations.

SAN FRANCISCO
 side span center span

I
— cable tower

MARIN
'— proposed line of Golden
Gate Bridge

r tension

anchorage
V compression force
to foundations
^Z
tower ^v

31
 Initially, the Golden Gate Bridge seemed to design
itself. After the line of the roadway was established,
the towers were located. They needed to be as close to
each other as possible to keep the center span as short
as possible. On the Marin side, the floor of the bay
descends rapidly; therefore, the tower was situated San Francisco Pylon —r-.
very close to the shore. Because the foundations of the 1100
San Francisco tower would need to be sunk 20 feet
into bedrock and divers could not work more than 100
feet down, that tower was situated where the water
was 80 feet deep. The length of the main span turned
out to be 4200 feet.
To avoid unnecessary and expensive underwater
—
work, the San Francisco pylon the support between
the ends of the side span and the approach road was—
placed on the closest dry land, about 1100 feet from
the tower. That distance was simply matched on the
other side to make the bridge symmetrical. The cable
anchorages could then be located behind the pylons
along the same axis.
The height of the bridge at the center of the
midspan (220 feet) and at the two towers (210 feet)
was determined by the U.S. Navy to ensure clearance
for its fleet. The approximate depth of the roadway
structure was to be 30 feet. The lowest point of the
main cable would be approximately 10 feet above that.
The optimum curvature of the cable for the most even
distribution of forces over a 4200-foot span would
require a sag of about 470 feet. Add these dimensions
to the height above the water and you have the
approximate height of the towers. Finally, a width of
90 feet was chosen for the road. A smaller width
would have seemed visually disproportionate for a
structure of this magnitude.
33
 CONCRETE SECTIONS
OF CABLE ANCHORAGE

Weight Block

%
Anchor Block

N :^

location of steel work

inside anchor block

Base Block
 Once the design of the bridge had been finaUzed,
work began on the anchorages. Each is composed of
three main pieces: the base block, which is keyed into
the bedrock; the anchor block itself, which is keyed
into the base block; and a weight block, which simply
rests on top of the anchor block. It is the tremendous
weight of the anchorage pushing down that counter-
acts the pull of the cable.The ends of the cables were
attached to a series of enormous eye bars that were in
turn fastened to heavy girders at the back of the
anchor block. The girders and the eye bars were to be
embedded in concrete.

35
 The two steel towers stand on massive concrete
piers. Because the San Francisco pier is 1100 feet off
shore, a jetty was first built for the delivery of
materials, equipment, and workers to the site. After
blasting away much of the rock at the end of the jetty,
the workers built an enormous elliptical concrete ring
called a fender. Its base went down 20 feet into the
bedrock and its rim stood 15
above the water.
feet
This not only served as a coffer dam during construc-
tion but also provided permanent protection for the
finished pier.
Because most of the concrete was needed only
beneath the two legs of each tower, the central section
could be left hollow. This space and the area between
the fender and the pier were eventually flooded for
added weight. Steel angle irons that extended some 50
feet into the concrete would eventually be tied to the
legs of the tower for additional stability.

cribbing

Site of Marin Pier

The pier for the Marin tower was to be built in

shallow water near the shore; therefore, construction
was less complicated. A coffer dam was first built

around the site. The part that extends into the water
is built with cribbing —heavy timber boxes that are
filled with rock and sunk into position. The ends of
the cribbing are connected to land with a simple rock
dike. The entire structure is then sheathed on the
outside with steel sheet piling —interlocking pieces of
steel that are driven into the ground. Once complete,
the area was pumped dry and the site was excavated
to expose a solid footing for the pier.
 The two towers of the bridge are identical. Each has
two legs connected with cross-bracing to stiffen the
structure against the considerable force of the wind.
The legs are built up from clusters of steel tubes
called cells. The cells are 3V2 feet square and approx-
imately 45 feet These are made of angles and
tall.

plates riveted together. To help distribute the weight

of the legs and prevent them from crushing the
concrete immediately below, they sit on thick plates
that are pinned in place with long dowels.
The cell clusters are lifted into place by a creeper
derrick (a truss bridge), on top of which two
stiff-legged derricks have been placed. Once the legs
of the towers reach a certain height, the creeper
derrick is hoisted up between them. Since the pier for
the Marin tower was a far simpler undertaking than
that of its counterpart across the gate, the Marin
tower was completed well ahead of its partner.
 The first bridges across were the temporary foot-
wire reel on bridges that extend all the way from anchorage to
unreeling machine anchorage. They were used by the workers in spin-
ning and preparing the main cables. The surface
of the footbridge is made of panels of redwood
planking that rest on rows of parallel steel ropes.
Approximately 3 feet above the footbridges is a guide
wire that has been carefully adjusted to hang with the
precise sag of the finished cables.
Each cable is made up of 61 separate strands; each
strand is spun from a continuous length of wire.

hauling rope

The which arrives at the site on large reels, is

wire,
initially fed through a counterweight system to
ensure an even tension before it passes over a large
wheel called a traveler.
The end of the wire is looped around a strand shoe
where it is temporarily clamped. The traveler is fixed
to a hauling rope. When the signal is given, the rope
starts moving, carrying the traveler and its cargo up
lydraulic jack to hold toward the top of the tower.
trand shoe in place On top of each leg is a specially designed structure
luring spinning operation s^. called a saddle. This is a precisely shaped seat with
grooves in which the finished cables will eventually
rest. Because of its tremendous weight, each of the
four saddles is lifted in three pieces.

end
temporarily
clamped
 \:

hauling rope

41
 sleeve wire (actual size)
yjj/JyJ.yJ/////>//yyjjT7yr
4
U. vmrTTTTTfTn^r
Every traveler first carried four wires. Eventually that is designed to prevent the wires from pulling
they carried six wires to hurry the pace of the apart. As the travelers move back and forth, the grow-
construction. Only two wires are shown here. ing strands are continually checked against the guide
Two separate travelers, each spinning a different wire to ensure that the exact curvature is maintained.
strand, move toward each other. When they meet in As work begins on a strand, the last pieces of eye
the middle of the center span, the wires are slipped bar are installed at each anchorage, linking those
from one wheel to the other. already embedded in concrete with a particular
Each traveler now returns to its starting point to strand shoe. Once a strand is completed, the shoe
complete the journey of the wire from the opposite (which has been held in a temporary position in front
side. of the eye bars so that loops can be wrapped around
Upon arrival, each loop of wire is removed and it) is pulled back and pinned in place between the eye
wrapped around its own strand shoe. bars.
The empty wheel then picks up a new loop from its
original reel and begins the journey all over again.
Each strand is made up of more than 400 lengths of
wire that lie side by side. That is about 500 miles of
wire per strand. When a spool runs out, a new wire is
connected to the end of the old with a tubular sleeve

completed
strand

steel pin

I— eye bar
43
 !!!!!i!!!ii!ii:isii;S:!!sss!:!:!:s!:::::
iiiiauiiiiiiiillti«B»iii>ii»»»**""'
lP'!:;s!:aiiiiiifiiiiiiBiiiiHSii>iai<:a!t
1
1111)
lllllll II !s:;iii:ssi!:!!!!iic!i
iBIiata SSSS'siSaitailiaafta'"
I aiiiaaaai«aaaiaii^>
1
II

•
aaaaiiaaaaaiiaaaiamii
aiaiaiaaaaniaaaiasi
aaiai(ai||||||l|l|(

•ifii
.<

One by one the finished strand shoes are pulled

^
i£S«IIM«l«ai(iSiiiai

back into their final position. As soon as the sixty-first

strand has been completed, the last of the concrete is
poured around the eye bars. The anchor block is
finished. Eventually, both pairs of anchorages will be
enclosed by high walls and a concrete roof
 Squeezing Machine

As the strands rise toward the towers, they first pass

through a large tube called a splay collar.
A squeezing machine is now fastened around each
cable near the tower tops. As it moves slowly down
a
cable, a series of plungers press and squeeze the
strands together. By the time the squeezing machine
has finished its job, the independent strands have
been transformed into a solid circular cable that is
greater than 3 feet in diameter.

splay collar

anchorage
lout walls and roof)
 cable band

A cableband is clamped every 50 feet along the

finished cable. Suspender ropes, each approximately
2 inches in diameter, are then hung over the cable
bands. These steel ropes support the road itself. Each
suspender rope has been cut to a specific predeter-
mined length to ensure that the finished road will
hang at exactly the correct height and camber.
All the suspender ropes are delivered to the site

with their ends embedded in steel cylinders called

sockets. Each rope enters a cone-shaped cavity within
a socket, where its individual wires are untwisted to
look something like a wire brush. The socket is then
inverted and filled with molten liquid zinc, which

permanently fuses rope and socket together.

Each cluster of four sockets disappears into a 25-
foot-high post. In fact, the post rests on the tops of the
sockets. It is neither bolted nor welded to the
—
sockets it simply rests on top of them. Each post is
connected to its counterpart on the opposite side of
the roadway by a deep floor beam. The floor beams
support smaller beams called stringers; stringers
eventually support the roadway itself The only
connections between the road and the cables are
those sockets.
socket

47
 To prevent a cable from undergoing concentrated
stress that could weaken it, its curvature should
remain constant. To help maintain the desired sag,
two 25-foot-deep stiffening trusses run along each
side of the roadway. A third truss, made of lattice
girders, lies in a horizontal position: it ties the stiff-
ening trusses together. The third truss is the wind
truss. It is designed to minimize bending caused by
wind pressure.
Construction of the roadway begins at the towers
and extends out over the water in both directions and
at the same pace to keep loading on the tower and
cables even. Sixmonths after workers began to hang
the roadway steel, the two halves of the main span
met in the middle.

rocker arm
Roadway Construction Sequence

1 1 1 1 1 1 1 1 rrnTiT]7T7-rri I . . , ....... ..i .

!• ' I I I I
1 I I I ., , , , r I

The would not achieve their intended sag

cables
until the entireroadway was in place, including the
concrete deck, railing, and streetlights. Only when
most of the weight was hanging did workers wrap the
cables with a layer of wire to help maintain their
shape and protect them from the elements.
In any large structure, engineers must pay
particular attention to the likelihood of movement
caused by high or gusting winds and extreme temper-
ature. Because it was intended for cars rather than
trains, the designers of the Golden Gate Bridge could
create a structure that would control this movement
rather than trying to prevent it. The ends of the road-
way assembly are actually hinged to the tower, so
the entire structure can rotate slightly on a
horizontal plane. The bottom corners of the two stiff-
ening trusses rest on flexible vertical columns called
rocker arms. In unusually hot weather the roadway
will expand, the camber will flatten out, and the ends
of the bridge will slide toward the towers. Expansion
joints are built into the road surface on both sides
of each tower and at the pylons so this movement
doesn't affect the cars above.

49
 In its day, this bridge created by Strauss and world. This title currently belongs to the Akashi
Clifford Paine was a masterpiece of engineering tech- Kaikyo bridge in Japan, shown below for comparison.
nology and building efficiency. Construction took less Not only are larger and larger suspension bridges
than five years. But while the Golden Gate may be the being built today, new ideas and technologies are
most famous bridge in the world, it has long since being explored. Because large bridges are expensive
relinquished its claim as the longest bridge in the undertakings, engineers continue to seek ways to
build them more efficiently. New aerodynamically shaped long strands to the site). While some towers are made
prefabricated roadways are replacing the older stiffening of concrete, heavy but very stiff, others are using

trusses, reducing construction time as well as the impact much than those of the Golden Gate Bridge.
less steel
of the wind. It is now possible to prefabricate cable Rather than simply restricting movement with thick
strands to eliminate air-spinning time (if there is a walls, builders are now installing sophisticated
vehicle large enough to transport the mile-and-a-half- devices inside their towers to dampen it.

51
 PONTE DE NORMANDIE
Honfleur. France. 1990 The French highway from the tower or towers. Unlike the main cable of a
administration decided to build a new bridge across suspension bridge, cable stays are straight and anchor
the river Seine with a clear span of 2800 feet. It directly into the roadway itself Each forms the third
would need to rise 165 feetabove the water so as not leg of a triangle. The cable stays are in tension while
to obstruct ship traffic They considered a suspension the road and tower are in compression.
bridge but eventually rejected the idea after looking
more closely at the geology of the area. Because there
was no solid rock into which anchorages could be tension
cast, each would have had to do the job entirely with
its own weight This would have required massive
and therefore prohibitively expensive structures. So
instead they settled on a kind of bridge that needs no
such anchorages. These are called cable-stayed
bridges, and at the time of its completion in 1994,
Ponte de Normandie was the longest one in the
world.
compression
In a cable-stayed bridge, the roadway is supported
by a series of cables extendmg in one or two planes
tf

cable stay

transverse cable ties to dampen

vibration of cable stays during
winds
 Once the foundations were in place, two 700- Concrete was used nearer the towers and between
and steel towers were then erected.
foot-higti concrete them and the approach roads to give the structure more
The roadway was built of prefabricated sections. rigidity. Steel was used for most of the main span because
Those of the side spans as well as that part of it is lighter yet still very strong.
the main span closest to the towers were made of
concrete. One section of concrete roadway was
cantilevered out from opposite sides of a tower simul-
taneously to maintain balance. As each new section
was one before it, two cable stays were
fixed to the
—
attached one on each side of the roadway.
Those sections making up the bulk of the main span
were steel. These were floated out and lifted into
place. The whether of
cross-section of the roadway,
concrete or steel, is basically a shallow box aerody-
namically designed to minimize wind resistance.

EINE
.^ , ^^> ,' " -;»—:» '
^3 J , —

53
Double-plane
cable-stayed
bridge
When cable stays are used in a single plane, they must
be anchored to the center of the roadway. Because
there is a greater likelihood of twisting in a roadway
not supported along both edges, the depth of that
roadway has be increased. In both cases, these are
to
bridges that use less material than a suspension
bridge would to span the same distance —primarily,
though not exclusively, by eliminating the anchor-
ages. This makes them comparatively economical,
and since World War II, cable-stayed bridges have
become increasingly common.
In order for a cable-stayed bridge to use as little

material as possible —the goal of all bridge engineers

— it must be continually checked and rechecked
during construction. This kind of precision is only
possible with the aid of the computer.
On the other hand, the longest cable-stayed bridge
in operation is still only half as long as the longest
suspension bridge, and this relationship isn't likely to
change. The amount of structure needed to withstand
the tremendous compression forces in the roadway of
a huge cable-stayed bridge would eventually make it

economically impractical.

There is little doubt that with the development

of new technologies and stronger materiails, bridge
builders will continue to push the limits — either
spanning greater distances or replacing existing
bridges with stronger yet lighter ones. But no matter
how sophisticated these magnificent structures may
become, we can count on the fact that they will
continue to reveal themselves to us as long as we
continue to look.

55
 Tunnels
If bridges are the most forthcoming examples of
engineering in this book, tunnels are their painfully
shy cousins. They live only to serve, and to serve they
must hide. Very few things about the way tunnels
look as we pass through them are even likely to catch
our attention. So while bridges, skyscrapers, domes,
and even a few dams enjoy varying amounts of
popularity, I think it's fairly safe to say that only an
engineer could love a tunnel.
Over the centuries, tunnels have been built for all
kinds of reasons, from burying mummified bulls and
delivering drinking water to mining salt and moving
people. But regardless of their different uses, the
shape or geometry of tunnels is quite similar. All
tunnels have weight on top of them, and most receive
pressure from the sides, which makes the arch the
ideal shape.
When a tunnel is bored through a mountain or
constructed under water, it may have to withstand
pressure from every direction, including below. In
this case, the arch is simply made continuous, so it

becomes more or less cylindrical.

The way tunnels are built depends primarily on
the technology available, the type and condition of
material through which they must pass, and their
ultimate length.The examples that follow have been
chosen not only to show various possibilities but to
illustrate the builders' level of determination and
inventiveness, about which the finished structures
say so While their initial costs may be high,
little.

well-designed and well-built tunnels can last indefi-

nitely and require relatively little maintenance all of —
which makes it that much easier to take these
remarkable and often indispensable structures
completely for granted.

57
 Two Ancient Tunnels

profile of landscape

Central Italy, A.D.41: To increase his land holdings in Some time in the sixth century B.C.on the Greek
the vicinity of Lake Fucinus, the emperor Claudius I island of Samos, a tunnel, 3400 feet long,had been cut
had his engineers drain the lake. This required the in similar conditions to carry water pipes. Perhaps to
planning and digging of a three-and-a-half-mile-long speed up construction, it was dug from both ends at
tunnel through soft limestone. Once the engineers the same time. Unfortunately, the two halves missed
had chosen a likely path for the tunnel, they created each other by about fifteen feet and had to be
a profile map of the terrain along its route. Using a connected with a sharp S curve. Over the following
leveling instrument called a chorobate as well as centuries, engineers began using closely spaced
measuring sticks and strings, they translated the vertical shafts along the line of a tunnel to avoid this
landscape into a set of precise but imaginary steps. At problem. If diggers didn't run into a shaft, they knew
specific intervals they recorded the distance between they were off course and could make adjustments.
the top of each step and the gi'ound. These vertical And since the engineers could determine the length of
measurements and the horizontal distances between each shaft precisely from the profile map, the shafts
them were then drawn out, creating an accurate helped establish the desired grade, which is particu-
picture of the mountain on which the tunnel and its larly important for tunnels meant to carry water.
portals could be located.

r— line of tunnel floor

_±
 /"

<^

An important consideration in all tunneling is some-

thing called stand-up time —the length of time an
excavated passage will stand without requiring any / ^^
kind of support. One advantage of tunneling through
rock can be its long stand-up time. If the rock is solid
and reasonably dry, a tunnel may not require support.
Although this happened to be the case with both the
Greek and the Roman tunnel, there is no guarantee
that even rock will behave itself once you start cutting
into it.

The obvious disadvantage to working in rock is that

it can be very difficult or at least time-consuming to
cut and remove. The hand chiseling required to make
ancient tunnels was supplemented by a method of
tunneling called fire quenching. The diggers build a
fire right up against the face of the tunnel. When the
vertical
rock gets very hot, they throw cold water against it.
shaft
The sudden change in temperature causes the rock to
split, making it easier to break up. But working

conditions in both tunnels must have been truly

horrific. Not only would they have been incredibly
hot, the diggers would have had to contend with
smoke, steam, and a variety of noxious gases. It is
recorded that 30,000 men excavated Claudius's ?.
tunnel for more than ten years. How many of them
actually lived to see it finished is not recorded.
tunnel

^
^i^'.^

^.
rv

1
n
<,
^.f

59
 Profile of Hoosac Mountain and tunnel

proposed line

Hoosic actual line as begun

River -
--4-
Plan View
Z proposed line

Hoosac Tunnel
North Adams, Massachusetts, 1855-1876: When the With assurance from the state geologist that the
Troy and Greenfield Railway Company was chartered mountain rock would behave in a uniform and pre-
in 1848, its goal was to create a link between Vermont, dictable way, have excellent stand-up time, and
Massachusetts, and the city of Troy, New York. present no serious water problems, the engineer
Unfortunately, the most efficient route for the new optimistically began his project. The tunnel was to be
line ran through a mountain range in north western arch shaped for additional stability, approximately
Massachusetts, specifically over a mountain called twenty feet wide and twenty-one feet high, and would
Hoosac. Because tracks can't be too steep or their carry a single four-and-a-half-mile-long track.
curves too abrupt, train travel over mountainous According to the chief engineer, it would take about
terrain always takes longer than it does over flatter four and a half years to build — less if the workers
ground. Tunneling, in spite of the difficulties dug a couple of shafts to increase the number of
involved, is often the best alternative. In fact, the idea headings, or locations from which the tunnel could be
of tunneling through Hoosac had already been excavated.
proposed for an earlier canal scheme and it soon The first two years were spent surveying the moun-
became the preferred choice of the railroad company. tain.Once an accurate profile had been produced, the
builders could locate the two entrances and establish
 Deerfield
River

the line between them. Since the two portals ended up By excavating what is called a top heading first, they
more or on the same level, the tunnel was to rise
less could determine whether or not the rock right above
slightly toward its center to ensure adequate the tunnel was stable. If not, they would support
drainage, if any water was encountered. Also, instead itwith heavy timbers as they went. Once the top
of approaching each other on the same line, both heading had been carried some distance into the
halves of the tunnel would enter the mountain at a mountain, other workers began removing the six-foot
slight horizontal angle, the idea being that when they layer of rock immediately below the floor of the top
eventually met, only a gentle curve would be required heading. This is called a bench. Eventually a second
to connect them. The engineer wasn't taking any six foot bench was excavated down to track level.
chances. The drastic S curve of Samos may have been Staggering construction in this way meant that not
fine for water pipes, but it would have caused a huge all the workers would be up against the same face
locomotive to derail, ripping up the tracks in the at the same time. At the east end, the contractor
process —which is not conducive to maintaining a decided to dig a bottom heading instead of a top
schedule. heading. Once the tunnelers had gone a certain
When construction finally began, the workers team working behind them b^gan
distance, a second
excavated both faces in sections. At the west end, they removing the upper layer of rock, called th^.^pe"'^
began by digging at the top or crown of the tunnel.

-=^
drilling the rock loading the powder

Like most hard-rock tunnels, Hoosac was to be dug In either case, the holder would rotate the drill

by the "drill and shoot" method. First, eight or ten slightly after each hit. Once several holeshad been
holes were drilled up to three feet into the rockface. drilled, they were filled with gunpowder. Everyone
This was done by one man with a drill and a hammer then got as far from the face as possible, except for the
or by a team in which one man would hold the drill man whose job it was to light the fuse. Presumably, he
while the others took turns hammering. was either the fastest or the newest member of the
team. When the smoke cleared, the group returned
and began breaking up and removing the shattered
rock, called spoil, in a process known as mucking. This
same tedious and dangerous sequence would be
repeated over and over until the job was done.
-^
running
 i

pneumatic drills on
movable carriage
(tracks not shown)

compressed
air hose —

Eventually a number of important changes put The rock at the east end was, as the geologist had
Hoosac literally at the cutting edge of hard-rock promised, solid — so solid, in fact, that a couple of
tunneling. Pneumatic drills, which had only recently steam operated boring machines were brought in to
been perfected, replaced the old hand-drill process. pick up the pace. Unfortunately, they both failed.
Several of them could be clamped to a movable The west end was a different story. Almost immedi-
carriage, which would be pushed right up against the ately diggers ran into a large area of rock that
face. The compressed air that ran them was piped in behaved more like soil. It was loose and porous. No
from either steam- or water- powered compressors at sooner was a space excavated than it would fill up
both entrances. Now the workers could drill holes with this "demoralized rock" and water. After six
three or even four times as deep in a fraction of the years of struggle, the project was way behind schedule
time. Powder was replaced by a fairly new invention and quickly heading over budget. The contractor was
called nitroglycerin. It was far more powerful, and replaced, new engineers were hired, and the state of
if delivered in a frozen state, relatively safe. An Massachusetts, hoping to protect its investment, took
electrical detonation system was also now available. over.
To these technological improvements was added the
center cut method-a sequence of drilling and shooting
—
from the center of the face out toward the sides and
the labor was divided into specialized groups so that
every worker no longer tried to do everything.

Center Cut Method

63
 —
Hoosac Mountain resurveyed
permanent survey marker

west portal

To speed things up even more, the new engineer The real challengewas to carry the line surveyed
took the daring step of putting the east and west along the top of the mountain accurately through it.
headings back on the same Hne rather than having Standing in front of a portal and looking though his
them enter the mountain at different angles. First, transit, the surveyor could locate the exact center
he resurveyed the entire path of the tunnel, line of the tunnel by using the marker immediately in
extending it up the sides of the hills facing each front of him and the one across the valley as a guide.
portal. All the surveying was done with a transit Once he had them in line, he simply flipped the
basically a powerful telescope mounted so that it can telescope vertically 180 degrees until he was looking
be rotated vertically and horizontally degree by straight into the tunnel. A worker inside held a plumb
degree. A compass below the telescope indicates the bob —a pointed weight on the end of a string — and
orientation, and several spirit levels keep the instru- someone else held a lantern behind it so the surveyor
ment horizontal. As the survey was being carried out, could see what he was doing. When the point of the
eight permanent markers were erected at key plumb bob was exactly in line with the crosshairs of
locations along the route. the transit, the workers secured it to a wooden stake
driven into the crown of the tunnel. Fifty feet or less
into the tunnel, a second plumb bob was hung. It too
was moved around until its point lined up exactly with
the point of the first one. As the tunnel slowly moved
into the mountain, so did the procession of plumb
bobs.
site line across valley

survey marker in front

of west portal

Telescop

compass
spirit level
 To help keep the tunnel on line and to increase the
number of working headings, two vertical shafts were
sunk from the top of the mountain. The first was dug
to the east of a fault that had slowed things down near
—
the west portal. The second this one over a thousand
feet deep —
was located more or less above the middle
of the tunnel.
Water was continuously being removed from the
tunnel. Because of the tremendous height of the
central shaft, it had to be pumped up in stages, until
it could finally be emptied into a small ravine near the
top. An elevator was also installed to the full depth of
the shaft to carry workers, equipment, and spoil.

65
 feet of
The finished tunnel required almost 8000
western half This lining
lining, most of it in the
courses of brick
varied in thickness from five to eight
was built up over a wooden centering. Where
the
and
ground was not firm enough to use conventional
walls of the
straight wall foundations, the curved
lining were supported on an
invert—basically a
continuation of the cylindrical shape. Permanent
sides of the
drains were cut into the rock along both
lining.
The last pieces of the project to be completed were
portal. Although permanent brick lining
the formal stone arches around each
don't really give much of
moderately impressive, they
the space between
a clue as to what it took to create
them. Building the tunnel was a twenty-one-year
actual construc-
struggle, with over fifteen years of
claiming two hundred lives and costing
more
tion,
than five times the original estimate.
have felt
While the tunnelers at Hoosac must often
water, when their
as if they were working under
underwater
tunnel finally opened, the first real
for thirty-five
tunnel had already been in operation
years.
wooden centering
 —

temporary west face

timber
supports 1

drain
sm^-^
Finished east portal
to the left is the hole '^i7-

begun by one of the /!-?.«-=

unsuccessful boring
machines.

67
 Rotherhythe

access shaft
high tide
X'.

low tide

Profile of river and proposed tunnel

 Thames Tunnel
London, England, 1825: The need for an additional
crossing between Rotherhithe and Wapping had been
steadily growing, but the idea of building a new
bridge must have seemed almost overwhelming. Not
only would it have to be able to open in some way so
that ships could pass, its construction would further

constrict an already congested river.

An engineer named Marc Brunei proposed instead
that a tunnel be built. It would contain two road-
ways, each passing through its own arch-shaped
passage, with both passages running through a single
masonry block. His suggestion was probably not
greeted with wild enthusiasm, since some twenty
years earlier, two Cornish miners had almost suc-
ceeded in digging a small timber-lined tunnel under
cross section of Brunei's tunnel
the Thames, only to have their efforts abruptly ended Between 42 and 76 feet below the riverbed there
by the unpredictable nature of the riverbed. A sudden —
appeared to be a layer of blue clay an ideal material
rush of quicksand and water had sent the workers for tunneling. It offered good stand-up time yet was
running for their lives and the tunnel's backers into soft enough to dig through relatively easily and was
a different line of business. more or less impervious to water. If the information
But Brunei was convinced that with the proper was correct and Brunei could keep his tunnel within
planning, his efforts could succeed. He would first the clay, things would progress well. Using the cross-
need to know exactly what was below the river so that section, he determined the depth at which his tunnel
he could select the best route for his tunnel as well as should be dug (62 feet) as well as its maximum height
the most appropriate method for digging it. A series (about 20 feet). Accessible from two shafts, one on
of borings were taken at regular intervals to retrieve each shore, the finished tunnel was to be roughly
samples of the riverbed. Combining this information 1200 feet long. Determining the path of the tunnel
with that gained by the earlier tunnelers, Brunei was was one thing, however. Figuring out how to build it
able to create an accurate cross-section. was something else.

access shaft

69
 completely
assembled
shield

To provide his workers with the necessary protec-

tion from both the uncertainties of underwater
digging and the tremendous water pressure, Brunei
first had to create a shield. The concept was
apparently based on (or at least inspired by) his
observations of the damage done to wood by ship-
worms. The scourge of the Royal Navy, these mollusks
used their shieldlike shells to bore holes through tim-
ber and then had the audacity to create a rigid lining
in the wood with material they excreted.
Brunei's shield was made up of twelve independent
frames, each 21 feet high, 3 feet wide, and 6 feet deep.
Together they would permit work on a face of approx-
imately 800 square feet. Each frame would contain

three workspaces, called cells, and each cell was just

large enough for one digger and one bricklayer. The
base of each frame rested on a pair of plates called
•- shoe shoes, which would distribute the tremendous weight
over a greater area to minimize sinking.
 ;

y^\

iron ring

shield
cross section foundations

sinking the Rotherhythe shaft

The frames were connected but were not rigidly of an iron ring. The 36 inches between the inner and
fixed together, so they could be moved somewhat outer walls was reinforced with iron rods and filled
independently. The top and sides of the shield were with rubble and cement. A second iron ring tied the
protected by heavy plates called staves. Between the tops of the walls together, and the whole thing was
shield and the tunnel face was a wall of short timbers capped by a wooden superstructure that supported a
known as breasting boards. Each board was held steam engine.
in place by a pair of screw jacks. Except for the As workers dug out the earth inside the shaft, the
breasting boards, the entire thing was made of cast thousand-ton structure began to sink under its own
iron. weight. The steam engine busily hoisted buckets of
To get the shield to its starting position, 62 feet dirt and pumped out water. Within three months, the
below ground, Brunei first built the Rotherhythe uppermost iron ring had disappeared below ground.
shaft from which the tunnel would be dug. Instead of At this point, 20 feet of earth below the bottom iron
simply excavating a shaft and then lining it, he built ring was carefully excavated so that bricklayers could
the lining abovegi^ound, like a gigantic circular build a permanent foundation. They left an opening
chimney. was 42 feet high, 50 feet in diameter, and
It 36 feet wide through which the shield would eventu-
made up of two concentric brick walls resting on top ally begin its journey.
71
 The tunneling itself was done in the following way.
A digger in each cell would loosen one pair of jacks,
remove a single breasting board, carve out about four
and a half inches of clay, and then immediately
replace the board and the jacks that held it in place.
The diggers repeated the process until the entire face
had been excavated. When all 800
in front of each cell
square feet of the face had been carved away, the
shield was pushed into it.
But this process was complicated and had to be done
one frame at a time. First the protective staves were
jacked forward into the clay. The tilt of each stave was
adjusted as necessary to make sure the shield was
staying on course. Then the ends of all the jacks
supporting the breasting boards in front of one frame
were slipped off their own frame and onto those
immediately adjacent. The feet of that particular
frame could now be lifted slightly, transferring much
of its weight to the adjacent frames. Finally, powerful
jacks pushing against the brick lining inched the
frame forward. Once every other frame had been
moved forward, the process was repeated for those
left behind. It was an incredibly cumbersome under-

taking, but on a good day, Brunei hoped to be able to

move forward about three feet.

73
 Unfortunately, he was not to be blessed with many Yet somehow Marc Brunei beat the odds. With the
good days. Even with the support of a string of sinking of the second shaft (which wasn't actually
capable engineers, including his talented son, begun until 1840, when it was clear the tunnel would
Isambard, it took nine years of actual building time succeed) and its eventual meeting with the shield, he
to dig a tunnel 1200 feet long. The borings had given proved that even rivers do not have to be impediments
the wrong impression about the depth and consisten- to tunnel building.
cy of the clay. The workers ran into areas of gravel, Although Brunei's shield now seems a bit like a
which has no stand-up time. There were numerous platoon of creaking Star Wars robots leaning against
floods. (One occurred half way across the river that each other for support as they inch their way nerv-
was so serious the shield was bricked up and the ously through the muck, it did the job; the revised
project abandoned for seven years.) Explosions of version he built after the resumption of work in 1835
methane gas ignited by the candles and lamps created was Not only was his thinking
especially effective.
terrifying flames. And there was plenty of foul air farsighted, his tunnel was so well built that it is now
from centuries of sewage, which continually made the part of London's subway system.
workers ill.

•"*-J
^"
•^
h

<:r

<
^ CJ
^>

'^ "1 -

<i
 r- worker's airlock
cast iron lining
Greathead shield

•— bulkhead

— material airlock
pneumatic jacks

A little less than thirty years after the completion of ever. Increasingly, the work inside them was being
the Thames tunnel, the engineers Peter Barlow
first mechanized. They were pushed by hydraulic rams
and James Henry Greathead, used a much more instead of screw jacks. A heavy wall or bulkhead was
compact version of Brunei's shield to dig a second often installed behind the shield, complete with air-
pedestrian tunnel under the same river. Theirs was locks so that the work area could be pressurized to
basically a tin can instead of a box, and the 8-foot-in- help keep the water out. He even came up with a way
diameter lining it left behind was made of cast iron to pump concrete into the narrow space between the
segments bolted together to form rings. Once again outside of the lining and the surrounding earth.
the shield was moved forward by jacks pressing Over the next hundred years, Brunei's vision and
against this lining. The project took less than a year. Greathead's engineering would unleash an army of
As the population of London grew and things on the manmade ship-worms boring tunnels in every
surface became increasingly crowded, transportation conceivable direction. They worked at previously
systems, particularly trains, were forced under- unheard-of depths and through all kinds of soft and
ground. Greathead continued to improve the waterlogged material to create the first and still one
tunneling shield (which now bore his name), and in of the busiest subway systems in the world, the
fact the whole tunneling operation. His shields were London Underground, appropriately known as "the
now up to 20 or so feet in diameter and stronger than tube."

75
 —

Holland Tunnel
New York/New Jersey, —
1920 1927: Although popular segments into place, the real breakthrough came in
from the beginning, the steam-and smoke-filled the development of a successful ventilation system
London Underground reached its true potential only one that set the precedent for tunnels to come.
with the arrival of much cleaner electric-powered His solution was to divide each tube into three
trains. When the engineer Clifford Holland undertook —
horizontal layers one to handle traffic, the other
the building of a mile-and-a-half-long automobile two to move air. Four large ventilation towers, two on
tunnel below the Hudson River, his problem was how each side of the river, would house eighty-four fans.
to eliminate all the exhaust fumes so that drivers Forty-two of them would blow clean air into the
wouldn't be passing out at 40 miles an hour. lowest space in each tunnel, forcing it through narrow
Once Holland had settled on the basic design of the slots along the curb and into the traffic area. The
tunnel — actually two separate tunnels running about other forty- two would extract all the dirty air from
50 feet apart through the riverbed — the method
silty the roadway through roof vents into the upper space
of construction was pretty much a given: pressurized and eventually out to the towers, where it would be
Greathead shields. Although these shields were very discharged. When traffic was at its peak, the air in
large ( 30 feet in diameter) and equipped with mecha- each tunnel could be completely changed every ninety
nized erector arms to fit each of the cast iron lining seconds.
Jersey City exhaust

Arrangement of ventilation towers

Exhaust air out

'
— clean air in
77
 The Channel Tunnel
The English Channel/La Manche, 1987-1994: After One of the reasons for this decision was evidence
centuries of distrust punctuated by prolonged from earlier tunneling attempts. Below the water, the
military confrontations, the French and the British two countries are linked by a layer of soft rock called
were finally united by a mutual dislike of seasickness. —
chalk marl a combination of chalk and clay, ideal for
The body of water that has separated the United tunneling. Not only is this material fairly easy to dig
Kingdom and France for the past 8000 years can be through, it has a good stand-up time and is more or
very rough, making ferry crossings sometimes less waterproof Extensive borings and sophisticated
unforgettable. But Britain's unshakable belief in the sound reconnaissance allowed geologists to create
importance of retaining what amounts to a moat for what they hoped would be an accurate picture of the
security reasons forced travelers into the air or kept various layers of material below the channel, which
them hanging over the rails as they crossed one of the the engineers could then use to select the best route.
world's busier shipping lanes. With the political and To control travel through the tunnel and to avoid
monetary unification of the European Community, the enormous ventilation problems that a 24-mile-
the time was right for the creation of some kind of long automobile tunnel would present, the engineers
fixed physical link between the two old rivals. A decided to build only a rail tunnel. Now, instead of
variety of proposals were submitted, including driving your car or truck onto a ferry, you would drive
tunnels, bridges, and combinations of the two. In the it onto one of two specially designed trains. Rain or

end, a tunnel proposal was selected. shine, the crossing would take a painless thirty-five
minutes from terminal to terminal, only twenty-six
minutes of which would actually be spent in the
tunnel. A third kind of train, called the Eurostar,

UNITED KINGDOM would carry passengers from the center of London to

the center of either Paris or Brussels in roughly three
hours.

ENGLISH CHANNEL
V

crossover cavern
Profile of channel and tunnel
 northern running tunnel (to France)

piston relief duct croissants

Miiiiii
service tunnel

djpy crumpets

southern running tunnel (to Britain)

The Channel Tunnel is really three tunnels that run service tunnel to keep out any smoke or fumes that
one another for most of the journey. The
parallel to might result from a fire in one of the main tunnels.
northernmost tunnel carries trains from Britain to All three tunnels are linked by cross-passages approx-
France, the southern one from France to Britain, and imately every 1200 feet. The two running tunnels are
a smaller service tunnel travels between them. Its also linked at 800-foot intervals by smaller tunnels
primary function is to provide access to the main called piston relief ducts, which permit the air that
"running" tunnels for periodic maintenance and to builds up ahead of a fast-moving train in one tunnel
serve as an escape tunnel if some kind of problem to pass harmlessly to the other.
arises. Higher air pressure is maintained in the

FRANCE

LA MANCHE

79
 ^ 1 '- - Vt"--- , ]-
i: 3:
concrete lining segments
on conveyor

The tunnels were to be dug by specially designed which installs the segments of tunnel lining.
tunnel boring machines, or TBMs. These are space- Extending another 800 feet behind the TBM is the
age versions of the Greathead shield and also leave service train. It delivers the lining segments, carries
behind them a completely lined cylindrical passage. away spoil and supplies clean air, compressed air,
They are highly sophisticated and almost completely water, and electrical power as well as sanitation, first
automated. At the front end of each TBM is a aid, and dining facilities- in short, whatever might be
rotating cutting head, which is thrust against the necessary to keep things moving.
rockface by a ring of hydraulic rams immediately The tunnels were begun from access shafts near
behind it. These rams also steer the cutting head. A each coast, down which the TBMs and all the other
second set of rams forces large gripper pads against equipment could be lowered. After undergoing final
the wall of the tunnel to provide a firm surface for the assembly, six of the eleven TBMs to be used began
thrust and steering rams to push against. Behind the their journeys towards each other, three from Britain,
gripper pads is the control room, from which the three from France. The other five began moving
driver of the TBM can monitor the machine's move- inland toward the future portals. The service tunnel
ment. A laser guidance system keeps it absolutely on was to be finished first and would serve as a kind of
course. The last piece of the TBM is the erector arm. advance party for the running tunnels.
 ^ripper pads nycirauiic inrusi ana steering ram

rotating
cutting head

hydraulic gripper ram

The largest of the rotating cutting heads, all of But even in the state-of-the-art Channel Tunnel,
which revolve two or three times a minute, was things did not go exactly as planned. All the prelimi-
almost 30 feet in diameter. Each head was studded nary exploration had suggested that the British TBMs
with chisel-shaped cutting teeth or inset with steel would have to contend only with dry conditions, and
disks, or had a combination of the two. As it slowly it was with this in mind that they were designed.

rotated, the cutting head carved a series of concentric Needless to say, it wasn't long before a large amount
rings of hills and valleys in the chalk marl. The of water began pouring through fissures in the
in
natural stresses in the rock caused the hills to split off chalk marl forcing the TBM at the British
end of the
as the valleys between them reached a certain depth. service tunnel to a complete standstill. Months were
The pieces of stone through spaces in the cutting
fell lost while a liquid cement called grout was pumped
head and onto the first part of the conveyor system into the cracks, then a space above the TBM was dug
that carried it all back to waiting spoil cars at the rear out and lined with steel plates and sprayed-on
of the service train. concrete called shotcrete. Only when this oversized
umbrella was in place and the flood stemmed could
work resume.

81
7 piston relief duet rp- castriron Kning sd^nents
^* concrete lining segments
* ^ i

key
jr —=— ^'
.

^mi -^
'

\^ i
'"-J
«

cooling
7 y/ z' »
pipes
94.'-'' —
''
•
[/

»
»
*

-^ L_ * ^^
•
^'"^>^

--^ r^
-^
;$^
\

=<N
.V A ^== J
'^ (^
U
iiJ'--7 -i^
> J 4

cross section
of running
tunnel
 '

^^
rj .V

—'*-~ — — ^ »»^ »— --
I
— probe
piuue
-^ -^ -."^--T. «r

ji.
^
^and-cut tuniiel

'Id
-<B- ^ I <

<! ii(0& ^M &

finished service tunnel

and the rings

All three tunnels are lined with rings, heads of the British TBM's, the engineers simply
are builtup of segments. The last piece to be inserted aimed them downward to dig their own graves. Once
into each ring, a wedge shape piece smaller than all all the ancillary equipment had been removed, the

the others, is called the key, and serves as a little holes were filled with concrete, and the French TBMs
reminder that these structures all members of the passed over them and into the British tunnels.
arch family. Most of the segments are made of rein- Getting the spoil out of any tunnel requires careful
forced concrete. Those around the cross-passage and planning from the beginning, as does finding a new
piston relief duct connections are made of cast iron. home for it. When the tunnel is thirty-two miles long,
In October 1990, when the service tunnels were a the amount of spoil generated makes this planning
little over 300 feet apart, the TBMs were stopped. To critical. The British built an enormous sea wall
make sure that the halves of the tunnel were in prop- enclosing a couple of artificial lagoons near the
er alignment, a two-inch-in- diameter probe was entrances to their access shafts. As the spoil came up
drilled through from the British side. Once it had through the shafts, it was dumped into the lagoons,
burrowed into the French excavation, an access forcing out the water. Eventually the spoil dried out,
passage was cut by hand. Then the entire section was making England slightly larger The French, having
enlarged to its final diameter, using smaller excava- more wet spoil to deal with, mixed it with water and
tion machines called road headers. pumped it into a specially created lake over half a mile
Six months later, the running tunnels also broke from the coast. Once their spoil had dried out, it was
through. This followed what appeared to be a rather planted with grass. Though approach may not
this
chivalrous gesture but was in fact simply a matter of have increased the size of France, it has added to its
economics. Rather than going to the trouble and oxygen supply.
expense of dismantling and removing the cutting

Frfench TBM
\
J,
rv > I I 1 I I t

British people I
I French people

^ ''^'l t
__

British
cutting h^ad

83
 " ^-;> *' r 5-^ ^.,
3L 1 1 * J Ir^

normal operation doors closed

British crossover cavern

doors open
partial operation

3i:
3 C

closed section
Cross section of typical crossover cavern

finished coircrete linin

cooling pipes
 \- doors closed south running tunnel

French crossover cavern service tunnel

doors open north running tunnel

To keep the trains running twenty-four hours a day, caverns were being built, workers installed measuring
even when part of the tunnel might be temporarily devices deep in the surrounding chalk that would
blocked, the engineers built two huge rooms called allow them to monitor ground conditions. If they
crossover caverns about a third of the way in from detected a problem, they could increase the thickness
each coast. In these spaces, the tracks were linked so of the lining or the length of the bolts.
that a train could cross from one running tunnel to Massive doors were installed inside the finished
the other in order to bypass the closed section. At the fire and to keep the
caverns to prevent the spread of
next crossing would move back to its original track,
it air in each running tunnel separate. They would be
at which time the train that had been waiting could opened only when the crossover system was required.
safely continue its journey. While this would certainly Work continued for two more years after all the
slow things down considerably, it also meant that tunnels had broken through. Miles of wiring were
under all but the most extreme circumstances, the run for the security systems, signaling, lighting, and
Channel Tunnel need never close completely. pumping equipment. Also installed were two large
During the excavation, the service tunnel was the pipes through which chilled water would be continu-
only link to the caverns, bringing in supplies and ally fed, to help reduce the heat created by the high-
speed trains. Everything was then tested and
retested, including the trains. By the end of 1993, the
Channel Tunnel was finished, and in May 1994 the
most expensive civil engineering project in history
was officially opened for business. But as everyone
knows, records are made to be broken.

I
— service tunnel

carrying away spoil. Before it was lined, each space

was approximately 500 feet long, 70 feet wide, and 50
feet high.When necessary, the chalk marl around the
openings was reinforced with shotcrete and 12- to 18-
foot-long steel dowels called rock bolts. As the
Prefabricated st^ei sectiow of
the Ted Williams tunnel
 The Big Dig
Boston, Massachusetts, 1985-: Like so many cities
these days, Boston is being strangled by its own
success. Traffic is often so congested that nothing
moves. In an attempt to improve the situation, work
began on the Central Artery/Tunnel project, better
known as the Big Dig. While much of it is directed at
replacing the old elevated expressway with a new
and wider highway directly below it, a number of
connections that have to be made in this extremely
complicated and busy site have required the building
of several interesting tunnels.
The Ted Williams Tunnel runs under Boston
Harbor and connects a couple of major highways with
the airport. With plenty of clearance between the
floor of the harbor and the ships above, engineers
were able to avoid the more expensive deep tunneling
approach, using instead something called sunken
tube design. The underwater portion of this tunnel is
made up of twelve prefabricated steel sections, each a
pair of 40-feet-in-diameter tubes roughly 300 feet
long. They were made in Baltimore and towed up to
Boston where all the rough concrete work was done
inside them. This included building supports for the
road, enclosures for the air-handling passages and
utilities, and a complete lining. One by one, the
finished sections were then towed out into the harbor,
filled with water, and lowered into a specially
prepared 50-foot-deep trench. Once anchored,
each section was pumped dry and opened up to the
adjoining section.

87
 Not far away, three other tunnels are under Two tunnels, both ten feet in diameter, are bored
construction. They are also prefabricated, although through the ground below the tracks and then filled to
this time not in Baltimore and not of steel. And about half their height with concrete. These will serve
instead of running below the harbor, these must pass as guides to prevent the main tunnel from dipping as
under busy railroad lines. The problem is to get them it slides over them. A thick concrete slab is poured at

built without causing any major disruptions to train

service or damage to the track beds.
The solution the engineers have chosen, called
jacking, has been around for a while, although it is

generally used to install underground pipes. Powerful

jacks push the pipes through the ground as the soil
inside them is removed. This ensures that the
surrounding soil pressure remains unchanged,
minimizing uneven settling. The process has been
used recently to jack tunnels, but not on this scale.
The tunnel fondly known as Ramp D is made up of
two hollow concrete boxes 80 feet wide, 30 feet high,
with a combined length of 150 feet.
A work pit slightly larger than the tunnel must be
excavated on one side of the tracks. It will be enclosed
by a high concrete retaining wall, which is built first.
As the trench in which the wall will be formed is
excavated, it is filled with a soupy substance called
slurry to prevent collapse. This keeps the pressure in
the trench the same as in the surrounding soil. Once
the trench reaches the required depth, concrete is

pumped in from the bottom up as the slurry is

pumped out. When the concrete has cured, the soil

enclosed by the wall is excavated.

slurry wall refrigerating pipes

work pit face

A ^— road' header
hydraulic jacks spools for
lubricated cables
Cross section of w^ork pit and Ramp D
 the bottom of the pit and the tunnel boxes themselves
are constructed on the slab. The portion of the
retaining wall between the tunnel and the face, is

eventually broken out. Roadheaders inside the box

then begin grinding away the clay, which is carted to
a waiting crane bucket by various pieces of earth-
moving equipment. Two rows of jacks, one between
the two boxes and the other between the back of the
spacer tubes pit and the outer tunnel box, push the whole thing

(added between rear into the space created by the roadheaders. Lubricated
wall and main jacks cables immediately above and below the boxes help
as tunnel progresses) them slide along.
89
 ^^
X <\ V
/^

. ^Ij^ves forVrout pipes

Grouting gallery *

As we've already seen, one of the biggest headaches Because of both the unpredictable nature of tunnel-
for tunnel engineers is caused by changing soil ing and the changes incorporated since the project
conditions along the route. However, it is now possible began, the Big Dig, like Hoosac, has suffered cost
in many situations to make the soil behave in a more overruns and numerous slips in schedule. But also
uniform and therefore predictable way. The clay into like Hoosac, the Big Dig is an ambitious and far-
which Ramp D was jacked was actually frozen first by sighted undertaking.
pumping refrigerant through pipes sunk between The surfaces of our cities have become so clogged
the tracks. with traffic trying to navigate worn-out roads which
When the time came to build a highway under are constantly being dug up for repair or to replace
the South Station subway line, things were more overburdened utility systems, that just getting
complicated. Engineers had to deal with four different through is almost as challenging as finding a place to
soil conditions and a lot of ground water. Two parallel park. Subway systems built over a hundred years ago
tunnels were bored below the subway. A precise
first to help address this very problem are now running at
pattern of angled holes was then drilled through the capacity. Yet the importance and appeal of cities
floor of these tunnels into which short pipes called doesn't seem to be lessening. The increase in popula-
sleeves were placed. These would serve as guides for tion that comes with each new office and apartment
the main pipes through which workers injected a building, shopping arcade and sports arena, simply
special chemical into the unstable soil below to adds to the burden. But that's the price of success.
improve stand-up time and help control the water. There is little doubt that more and more tunnels
Once the soil had been stabilized, more tunnels were will have to be built just to keep our cities livable. And
bored below each grouting gallery, one directly below they will have to be dug through or below the maze of
the other. As each was finished it was filled with foundations, subways, and utility systems on which
concrete. A third row of tunnels, this one horizontal, we already depend. Complex projects like the Channel
was then bored between the grouting galleries. These Tunnel and the Big Dig remind us that engineering
would house huge concrete beams. When all the technology and ingenuity are up to the challenge. It's
pieces of this enormous subterranean log cabin were just a matter of cost. How much are we willing to pay
in place, the soil inside could be safely excavated and to help insure the health and success of our cities for
the four-lane highway tunnel built. generations to come?
 '.^H

#?»»>

It5

1* -
-& -

--- i.^
Subway / ,-

77
^
( \ ^^ grouting galleries

» »

.IE

./

^;-^'
; "^iii % 3 E
^
v) ^-i

.V.

'>'
^ stacked tunnels
new road tunnel
stabilized ground
w.

u*fO .

11 I

\ \
 —

Dams
Of all the big things in this book, dams feel to me
like the biggest. The highest ones would barely reach
halfway up the tallest skyscrapers, and with only rare
exceptions do dams stretch as far as the longest
bridges. Yet they seem bigger than both. Perhaps it's
because we find them off the beaten path, where
there is littlecompare them to. Or maybe it's their
to
simplicity. There aren't a lot of small pieces to dams
that might reduce their visual scale. Then again, it
might not be the way they look at all but, rather, what
they do. With a single bold gesture, dams affect every-
thing immediately around them and can have an
impact on life hundreds of miles away.
Regardless of their size, all dams have two
fundamental components, an impervious barrier to
prevent the flow of water and some kind of structure
to keep that barrier in place. When designing a dam,
engineers work with two primary elements. The first
is the shape and configuration of the structure. The

second is the material with which it is built. While the

four examples that follow present aspects of design
and construction common to all dam projects, they
also illustrate how the specific requirements of each
dam combine with the uniqueness of a particular site
to help engineers arrive at the most appropriate
design.
When all is said and done, dams have one job
controlling water. By raising a river's level or
diverting it, they help produce electricity, reduce
flooding, improve irrigation and navigation, and even
promote recreation. Controlling water, however, is
not as easy as it might sound, especially when you're
talking about millions of gallons. As every dam
engineer knows, water also has one job, and that is to
get past anything in its way.

93
 Ita Dam
Uruguai River, between Santa Catarina and Rio or brick, whereas embankment dams are made
Grande do Sul, Brazil, 1996-2000: In 1987, after primarily of rock, sand, earth, and clay. Because of the
several years of study, the government of Brazil high cost of working entirely in concrete and the
formally approved an ambitious dam-building project almost unlimited availability of sound rock in the
intended to meet the country's electricity needs well area, Ita was an embankment dam.
to be
into the twenty-first century. Of the twelve dams A straight wall across any river will soon be pushed
planned, the largest was to be built on the Uruguai over if it isn't heavy enough to resist the horizontal
River.The site was chosen because of the large force of the water behind it. The builders can either
amount of rainfall in the region, the expected height make it thicker or pile up material in front of it with
of the water at the dam and the potential capacity of enough weight to counteract the push of the water. If
the reservoir behind it, and the solid rock layer below the wall is supported at an angle, some of the water's
the soil. force will push downward and actually help keep the
Masonry dams are those built of concrete, cut stone, dam in place.

I / ^^^;^>.->_^-^' / ' spjrllway #1"

/ / y^ ^-^—f- futile Reservoir '

( ( / /^ \ f / ^^^ /

/ /
4— i:
This is the principle behind all embankment dams.
The impervious barrier at Ita would be a layer of
concrete. The structure supporting it would be a man-
made mountain of quarried rock about 400 feet high
and half a mile long.
As is generally the case, Ita dam was tobe just one
part — albeit a very big part — of a complex of
structures. It would also require two cofferdams,
three saddle dams (to seal off depressions in the land-
scape that fall below the expected height of the
reservoir), two spillways, ten tunnels, and a power-
house containing five electrical generators.

force of
water
weight of
support

wall with support

95
 O-V- _—

,K:

>-w:.

(»i

\ «^
*^v

<ii^h
»- -Vt ^/.
\^^s
c
'/ /^7/

>^' .**'^r^j

setting up the wooden templates

diversion tunnels — power intake

spillway #2

diversion tunnels
 To help and dimensions of each
establish the layout Once all the hills and valleys were in place, the
structure and to get some idea of how the water would pieces of the dam complex were added. By flooding the
behave, engineers built a huge scale model. Working model, the engineers could now begin to see how the
from detailed maps, they drew the path of the river on design was working. Readings taken from various
the concrete floor of a large, open building. Using a gauges were interpreted and analyzed, and the
transit for accuracy, they then set up precisely necessary adjustments were made. Eventually things
measured and cut wooden templates along the out- learned from this process were added to the
lines of the river. By filling the space between the data —geological reports, technical calculations, and
templates with crushed stone and cement, they were so on — required before a final plan could be
able to replicate the topography of the area. developed.

main dam
 Even as the model was being constructed and
studied, preparations for the project were underway.
Many people operated small farms along the river-
banks upstream from the dam site. Still others lived
and worked in the town of Ita. All of them would have
to be relocated, because by 2001, the only residents of
this valley would be fish. By 1990, many of these
people were settling into brand-new homes along the
brand-new streets of Nova Ita (paved with cobble-
stones from Old Ita). They had a new community
center, a small museum in which pictures and
souvenirs of the old town were displayed, and a new
church. Near it was the new cemetery, to which the

remains of Ita's deceased had been carefully trans-

ported. Three hundred and fifty miles of new roads
and ten new bridges were also created to carry
workers, equipment, and materials to the dam site.
But just when everything appeared to be ready, public
funds ran out and the whole project ground to a halt.
Six years passed before a consortium of public and
private partners was finally able to begin construction
in 1996.
Building anything in a strong current is difficult

and dangerous, so before construction of the dam

began, engineers diverted the river around the site.

Since the Uruguai River doubles back on itself,

making a large loop, they decided to connect the

waters above and below the dam site with five tunnels
driven straight through the rock. A large cofferdam
was builtwhich forced the river into the tunnels. A
second cofferdam, built downstream from the site,
prevented the river from flowing back into the work
area. Stone removed during the excavation of the
— —
tunnels in this case a hard basalt was used to build
both cofferdams and part of the main dam. All three
dams were built up in layers, a process that was
begun during the dry season. Although the ends of
the main dam were already being built up against the
banks, they would not be extended across the river
until both cofferdams were finished, the area between
them had been pumped dry, and the riverbed had
- expected height
been cleared to provide a strong rock foundation.
of reservoir
upstream
cofferdam
 cross section of Ita Dam

downstream cofferdam
(becomes part of main dam)

99
 If water can't get through a dam, it will try to go tremendous force of the water
to evenly distribute the
around or under it. For this reason a reliable water- pressing against the thin concrete face (18" thick at
tight seal between the concrete face and both the rock the bottom and only 12" at the top). As each layer of
foundation and abutments is essential. A trench is cut transitional material was placed, a cement curb was
into the stone along this entire line and a thick precisely extruded along the entire length of the face
concrete pad called a plinth is built into it. Holes are so that the finished surface would be as uniform as
drilled at regular intervals through the plinth and possible.
into the rock, and grout is then pumped down into Only when the transitional material and the plinth
them to fill any cracks. Long steel rods are also were in place (this took about two years) could the
inserted to tie the plinth and surrounding rock face itself be constructed. It was built in a series
together. of vertical sections each about 40 feet wide and
While the plinth was under construction, the dam extending the full height of the dam. Each section was
itself was being built up one horizontal layer of rock- enclosed between temporary wooden forms, and
fill at a time. These layers varied from three to six feet workers then inserted a grid of reinforcing steel. The
thick and were compacted using heavy vibrating first concrete was poured at the base, creating the all-

rollers. When the plinth was completed, layers of important watertight connection between the face
crushed stone only a foot and a half thick were built and the plinth. A flexible joint between the two would
up behind it. Known as the transitional area, this part permit the face to move slightly as the reservoir
of the dam must be built very carefully, since its job is eventually rose against it.

I \ ' ' ' « I

,
I
. •
*
( .
I'll
dn upstreaiq sjd^
'
crbssl sfectipn 6f dfeim Jqasi^

)0
 reinforcing
bars -

" ''
-^ wooden
'"'''-_
« -form
 The concrete was delivered down long chutes that
extended all the way from the crest. After each pour,
a steel platform the width of the section was slowly
winched up over the wet concrete. From the platform,
some of the workers vibrated the concrete to ensure
even distribution while others troweled it smooth. A
perforated pipe, dragged along behind the platform,
provided a continuous trickle of water to keep the
concrete from diying too quickly and cracking.
^- -r"
103
profile of spillway #1

plunge pool

Once the and the surrounding pKnth were

face
finished, the only place left for the
water to go was
over the top. For any dam, overtopping, as it is
known, can be dangerous. For an embankment dam it
can be disastrous. To prevent overtopping, all dams
are accompanied by one or more large concrete chutes
called spillways. These carry water from above the
dam to the river below it. At Ita, there are two of
them.
The top of each spillway isalways several feet below
the crest of the dam so that when the water in the
reservoir is unusually high,
it can be safely discharged

without threatening the dam itself. A row of curved

steel gates above both spillways controls the amount
of water that is actually released. To prevent erosion
at thebottom of the spillways, the chutes are curved
upward, sending the water into the air It then falls
harmlessly into huge pits called plunge pools before
resuming its journey downstream.

105
 The last major piece of the complex, and the
primary reason for the dam in the first place, is the
powerhouse. This is the building that houses the
generating equipment, and it is always built as far
below the level of the reservoir as possible to insure
that the water has as much force as possible when it
arrives.
At Ita, the water begins its journey to the power-
house by passing through another set of enormous
gates. Unlike the gates above the spillways, these
slide up and down like the portcullises of medieval
castles. The water enters one of five separate tunnels,
called penstocks, each over 20 feet in diameter.
Screens called either trashracks or stoplogs,
depending on the sizes of their openings, prevent
debris from passing through the gates that might
either block the penstocks or damage the turbines
below. Because of the tremendous water pressure that
builds up inside them, the penstocks are lined with
either concrete or steel.

?^^
gantry crane
to raise and
lower gates,
trashracks
and stoplogs
 penstock
turbine

cross section of power intake

Once reaches the powerhouse, the water flows

it

from the penstocks into a spirahng steel pipe called a

scroUcase. This in turn is wrapped around a bladed
wheel called a turbine. As the water leaves the scroll-
case, itpushes against the blades and spins the
turbine. The axle or shaft of the turbine connects
directly to the shaft of a second wheel called a rotor,
which is lined with magnets. The rotor turns inside a
large fixed rim called a stator. The movement of the

magnets transforms a small current in the stator into

a much more powerful current, which is then drawn
off and readied for distribution.
Because of the tremendous expense of such projects,
it is essential that they be completed
and producing
electricity as quickly as possible. For this reason, the
various parts of the Ita complex were built simultane-
ously. In June of 2000, the first electrical power
from
the Ita dam reached its Brazilian customers.
 ]

supports —

wicket gate
(adjustable)

stay gate
(fixed)

draft tube
(returns water
to river)
Turbine generator
109
 Hoover Dam
The Colorado between Nevada and Arizona,
River, The Bureau of Reclamation had been looking at
1931-1936: A dam across the Coloradowas proposed various dam sites along the Colorado almost since the
for four reasons. The first was to irrigate the arid turn of the century. By 1928 it had whittled the

Southwest. The second was to control the often number of possibilities down —
two Boulder Canyon
to
unpredictable river and minimize flooding. The third and Black Canyon. Both sites had the capacity for
was to collect silt carried by theand the fourth
river, very large reservoirs, but in the end, Black Canyon
was growing
to generate electricity. It wasn't until the was chosen. Its walls were higher and the river was

cities of southern California agreed to buy much of narrower there which meant that a smaller dam could
that electricity that the Hoover Dam project finally do the job. (It is hard to imagine that a solid concrete
got under way. structure 700 feet high, 1200 feet wide at its crest,
and 660 feet thick at its base could be considered
smaller than anything, but that's the dam business.
The river was first diverted around the site through
four tunnels. Each was about three quarters of a mile
long, 56 feet in diameter, and lined with a three-foot-
thick layer of concrete. By the end of 1932, the two
tunnels on the Arizona side were ready, and the
barriers blocking their portals were blasted away
allowing the water to enter.

Ill
 Immediately, workers began building the upstream impervious surface and its supporting structure one
cofferdam, which would eventually reach a height of and the same. In cross section. Hoover is basically a
98 feet and be faced with concrete. Within five right triangle with most of its mass near the base,
months, both the upstream and downstream coffer- where the water pressure behind it is greatest. The
dams were finished and the space between them was upper half of the dam curves in the upstream
pumped dry. direction. This portion also works like an arch,
Forty feet of accumulated silt and mud were then directing some of the force pushing against it into the
excavated to expose and prepare the bedrock. Much of sturdy canyon walls on either side. A channel was cut
the muck was hoisted from the riverbed by a into each abutment to support the ends of the arch.
specially designed cable system that linked both walls Although Hoover is an arched gravity dam, it is
of the canyon. The cables were suspended from highly unlikely, given its enormous weight, that the
towers which could be rolled along tracks in order to arch was really necessary. But people seem instinc-
pick up and deposit their loads wherever needed. At tively to trust arches, and just seeing that curved
the same time that the riverbed was being prepared, crest pushing back against the water makes anyone
the walls of the canyon were being cleared of all loose standing on the dam feel secure.
and unstable rock. This was one of the most danger- By the middle of 1933, construction of the dam itself
ous jobs on the project and was performed by men began. Because of the enormous amount of concrete
known as high scalers. Like performers in some kind that would be needed and to guarantee an uninter-
of demented circus, they hung from the tops of the rupted supply, two concrete mixing plants were built
canyon walls on long cables and operated powerful right at the site. A small railway line was also built
—
jackhammers all without a net. along the Nevada canyon wall to transport the
The Hoover Dam, like the Ita Dam, is basically a concrete from the downstream plant to the hoists of
gravity dam, but because of its shape and height it the cable system.
had to be built entirely of concrete, making the

cable system towers and temporary railway line on the Nevada side

L2
cross section of Hoover Dam
113
 ^
4^^-

iM

pipies of
cooling system
 The dam is composed of 230 vertical columns, each
approximately 25 by 60 feet in area. Each column was
builtup 5 feet at a time, and all were staggered so that
the wooden formwork that held the wet concrete in
place could be maneuvered more easily. The sides of
the columns were ribbed, either horizontally or
vertically, so they would interlock. This ensured that
the finished dam would act as a monolithic mass. By
limiting the size of each pour orworkers could lift,

make sure that the concrete spread evenly and

completely within the form. A number of vertical
shafts and horizontal galleries were built into the
dam as it rose. These would be used for inspection and
drainage, and also for grouting as the concrete dried.
Concrete is a mixture of crushed stone, sand, and
cement. When water is added, the cement undergoes
a chemical reaction called hydration. This results in
the formation of crystals that tie themselves and the
rest of the ingredients together. But hydration also
produces considerable heat. If this heat dissipates too
quickly or unevenly, cracks can form. Cracks in a dam
are not a particularly welcome addition, and the
builders make every effort to eliminate them.
The amount of heat that would be generated by and
then trapped inside the enormous amount of concrete
created two problems for the Hoover Dam engineers.
The first was the danger of cracking because of
uneven cooling as the heat slowly worked its way out.
The second was the fact that the joints between the
columns could not be grouted until the entire struc-
ture had cooled down. Without help, the Hoover Dam
wouldn't be cool enough for another 50 years.
still

To accelerate and control the cooling process, the

engineers embedded pipes in each lift and pumped
ice-coldwater through them. The rate of cooling was
constantly monitored and if necessary adjusted by

changing the temperature of the water. An 8-foot-

wide slot was left open at the center line of the dam to
house the large pipes that carried water between the
structure and the refrigeration plant. The slot was
filled in a series of 50-foot lifts as the dam grew
around it. The entire structure was successfully
cooled within twenty months of the last pouring early
in 1935.

115
LAKE MEAD

plug in
diversion
tunnel

S
route 93
 Immediately upstream of the dam are four thirty- buildings. The intake towers stand on ledges cut into
four-story-high intake towers. Water from the the canyon walls over 300 feet above the riverbed and
reservoir passes through them into the steel header can be closed by enormous cylindrical gates. The
pipes and penstocks, which lead either to the twin space between the base of the towers and the riverbed
powerhouses at the foot of the dam or to two outlet provides the pocket needed to collect the silt. The
primary function of the outlet buildings is to dis-
charge the water needed for irrigation downstream.

The major pieces of the complex are the two 500-

last
foot-long spillways —
basically massive concrete bath-
tubs through which water can be drained if the
reservoir gets too high. Both the spillways and the
intake pipes from two of the towers empty into the
diversion tunnels, which are plugged just above the
connection.

117
 Aswan High Dam
The Nile River, Egypt, 1960-1971: Twenty-four years
after the completion of the Hoover Dam, dam
engineers were once again eyeing a river for retrain-
ing—this one a lot wider and a lot more famous than
the Colorado. For more than four thousand years, the
Nile had provided the farmers along banks with
its

rich, nutrient-filled silt to replenish the fields and

water to irrigate them, and in so doing had helped
create one of the world's great civilizations. But the
Nile, like the Colorado, could also be unpredictable.
Some years it was too generous, flooding villages and
destroying crops. Other years it could be incredibly
stingy, causing widespread famine.
By the end of the nineteenth century,
population was steadily growing, particularly in and
around Cairo, the country's almost complete
when Egypt's
^^'^••^^ V/M^
dependence on the whims of the Nile had become
intolerable. In an attempt to address the situation,
British engineers designed and oversaw the construc-
tion of the first dam across the river, an impressive reduce flooding and guarantee a constant water
masonry structure at a place called Aswan. In the supply, making farming a year-round rather than sea-
years that followed, its height was increased a couple sonal operation. Egypt, although still dependent on
of times, but the dam still wasn't big enough to the Nile, would at last be the master of its own fate.

satisfy demands. Once again cofferdams were built above and below
Starting in the 1950s, engineers made studies and the site and the river was diverted around them. This
eventually drew up plans for a much larger structure, time, however, while the current between the coffer-
which, to distinguish it from its predecessor, was to dams was eliminated, the water was not. Because the
be called the Aswan High Dam. It would provide Nile wide and its abutments are low, the engineers
is

much more electrical power than the old dam both for settled on an embankment dam. Its impervious
industry and to improve Egypt's standard of living. barrier would be a solid clay core held in place by
By impounding a huge reservoir, it would also greatly layers of rockfill and sand.

cross section of Aswan High Dam

119
creating the grout curtain
 Although the Nile does not have the high abutments
of the Colorado, it does have a canyon. Unfortunately,

though, this one goes down rather than up,

descending roughly 600 feet below the river, and is
filled with sand and gravel. While the clay core would

keep water from passing through and around the dam

and an emergency spillway would prevent overtop-
ping, the engineers needed to provide some way to
keep the water from escaping under the dam,
threatening it from below.
They solved the problem by building an underwater
barrier called a grout curtain —basically, a continuous
wall of interconnected columns. A predetermined
pattern of holes was drilled across the river. The
deepest ones went down the full 600 feet to bedrock.
As each hole was drilled, a three-inch-in-diameter
pipe was inserted into it. Once a hole reached the
desired depth, the pipe was pulled out as grout was
forced in. As soon as the grout passed through the end
of the pipe, it quickly expanded into the surrounding
material. When this dense mixture of grouted sand
and gravel set, it created a permanent column almost
As rows of columns were drilled
five feet in diameter.
and grouted, the workers eventually built a contin-
uous waterproof barrier.
By the time the various layers of compacted
material were in place above the grout curtain, both
cofferdams had been incorporated into a structure
that now stretched more than half a mile up and
down the riverbed at a height of 400 feet. The Aswan
High Dam took 30,000 workers almost ten years to
complete and claimed 500 lives in the process.

121
 Itaipu Dam
The Parana River between and Paraguay, 1975-
Brazil
1991: In April of 1974, the governments of Brazil and
Paraguay signed a treaty that set in motion the
building of the world's largest hydroelectric dam. The
agreement followed a four-year study during which
various locations and combinations of dams were
considered. In the end, the countries decided to build
a single large dam at a place called Itaipu. With its

eighteen enormous turbines, the finished dam would

provide Paraguay with more electricity than it could
use and satisfy about 30 percent of Brazil's current
needs.
The structure, almost five miles long, is a combi-
nation of both masonry and embankment dams. The
first piece of the dam to be built is a solid gravity
concrete structure through which the entire river was
initially diverted. Next to it, extending across the
river, is a hollow gravity concrete dam. Since the
hollow part doesn't have to be as heavy, the sloping
faces are supported by a row of parallel walls
separated by enormous cavities. Extending from both
ends of this structure are two concrete buttress dams.
These have only one face, and it is supported by a row
of exposed sloping walls. Embankment dams with
clay cores and outer shells of rockfill or earth diversion channel while the solid gravity structure
complete the structure. In fact, the only significant was being built. As soon as it was finished, these dams
dam type not represented at Itaipu is the arched dam, were blown up, and the course of the river was
but two of them were used originally to block the diverted over the rubble.

hollow gravity solid gravity buttress

dam , r-rdam — 1 i^ dam
buttress
dam
arthfill
am

spillway

:=^j^
Understanding just how big a dam is can be difficult,
particularly one like Itaipu. Although you can take it
all in from the air, you still can't really feel the extra-

^prdinary scale.

123
\
Wi:i. f V ItlllT
So here is the view from what is called the upstream

road of some of the 34-foot-in-diameter penstocks as

they emerge below the crest of the dam and begin
their descent to the turbines. That's an average-sized
compact car parked up against the second penstock.
bedrock
 Now, when we take the previous image and insert it
into its rightful place in a cross-section of the main
dam and powerhouse, perhaps it is a little easier to
comprehend the reality of building big.

cross section of Itaipu Dam

127
 i

-^

\y\

1*.

\
 Domes
The world's great domes have the abihty to move
us, both hterally and figuratively. On the one hand,
we find ourselves spinning around just to take them
in. On the other, they elevate our spirits by drawing

our gaze heavenward or by appearing to float weight-

lessly. But the builders of these remarkable struc-
tures would be the first to remind us that defying

gravity isn't something to undertake lightly. Domes

are, after all, roofs. No matter how big or impressive
they may be, they are up there for a reason, and up
there they must stay.
The and purest domes were single masonry
earliest
vaults, and was more or less
their interior shape
reflected in their exterior lines. Over the centuries,
however, domes began to serve as beacons, and as
such they were built higher and higher above the
ground. In time, the proportions of the interior shape
were no longer satisfying from below, so architects
solved the problem by building a second dome within
the first. Eventually, as many as three domes or
domelike shapes were stacked one above the other. As
the complexity of these structures continued to grow,
so did their importance as symbols of religious,
cultural, and government institutions.
During the twentieth century, as technology devel-
oped and materials grew stronger, enormous domes
were built to cover sports arenas and meeting halls.
By the end of the century, the word "dome" had
become so synonymous with large gathering places
that it was used to refer to a number of vast spaces
covered with curved roofs of various kinds. "Domes"
such as the Georgia Dome, the Millennium Dome,
and the SkyDome are certainly considerable feats of
engineering, but they are just too big to inspire. They
may impress us, even overwhelm us, but transport us
beyond ourselves, like the Pantheon or the dome of
St. Peter's in Rome? I think not.

129
 occulus

portico —
 The Pantheon
Rome, Italy, 118-125: Upon his ascension to the by a 27-foot-in-diameter eye, or occulus, right at the
throne, Hadrian, the thirteenth ruler of the Roman top. It was a Not only would a very
brilliant design.
Empire, inaugurated something of a public relations large audience be able to see Hadrian conducting
campaign. As was traditional, he forgave certain debts official business, but he would do it beneath the gaze

and provided elaborate and generally gruesome enter- and therefore with the implied support of the gods
tainments at the Colosseum, but he also wanted to themselves.
create something for which he would be remembered. Regardless of whether domes are hemispherical,
As a gifted amateur architect, Hadrian understood egg-shaped, or saucer like, all them obey
of certain
—
how much power a building could have especially a fundamental principles. To understand how a dome
big building. Knowing also that even an emperor can works look at the one below which is covered with a
use a few allies beyond his earthly realm, he worked grid. The vertical lines are called meridians. The
—
with his architects to design a pantheon a temple to horizontal lines are called parallels. At the top of the
all the gods —to replace Agrippa's old pantheon, dome, as in an arch, the meridians lean toward the
in
which was now in a state of considerable disrepair. center, forcing the parallels into compression. At the
The building they designed had two main parts. The bottom, the meridians push outward, stretching the
first was the kind of entryway people would expect in parallels and putting them into tension. At a partic-
—
front of a temple a raised portico, complete with ular point between the two sets of parallels is an area
columns and a triangular pediment. The second, that is neither in compression nor in tension. This is
however, was to be something of a surprise a vast — represented by the dotted line. If a dome is to stand,
circular room covered entirely by a single dome. This the tension and compression forces must both be
enclosed but unobstructed space beneath a manmade safely handled.
concrete sky would be linked directly to the heavens

parallel '/^
N^ meridian
/

<^'^.^

131
^ ^
^3 c
?=>?*

-==
H 1 i^

^^ ^
k^
4

>\ ^5
1 ^
i ^-j J^S^^vm

i 4 ^sfc^^V
-
1

=4=

^^ fl ^^
1
^: > 'n
^.,^
^^
iS
P^,.
£4
H;5~^
\
i
i
S^
^^%y

w:^ S'
I al1
.
^
'

2bh

'^^ "^J
'^p '^^

1
^^H^H
-^^^^H^^^^^p JK VV^
-^^S^S^^^^B ^ij
f^-^mM^^ 2r^'A<"-

VmhHhS t^MHl^^^H
'

Ss^A

1
 Though the Pantheon was to be the largest concrete
dome in the world, Roman builders were not inexpe-
rienced with either the shape or the material. Its form
was a logical extension of the vault, which in turn had
been a logical extension of the arch. The brick-faced
concrete cylinder upon which the dome would rest
was actually a ring of massive piers, linked by
relatively thin walls and large buried arches. These
arches were designed to carry the tremendous weight
above them and channel it around windows, passages,
and niches to the piers and foundations below.
Although the Pantheon was to be made almost
entirely of concrete, wood played a very important
role in its construction. Workers filled the space that
would eventually be enclosed by the cylinder with a
huge timber scaffold. This manmade forest provided
platforms for the wall builders and supported the
hemispherical form over which the concrete dome
would be poured. Five horizontal rings of trapezoidal
indentations called coffers were to be cast into the
interior surface of the dome. These would not only
add to its visual complexity, they would also reduce its
weight. These shapes first had to be built as projec-
tions on the surface of the form.
Concrete is created by mixing aggregate, such as
stone and sand, with cement and water. To further
reduce the weight of their dome, the builders of the
Pantheon used different types of aggregate in its
construction. For instance, the material at the base of
the dome is a heavy stone called basalt, while at the
top it is a much lighter volcanic stone called pumice.
Around the base of the dome, several layers of
concrete were added to counteract the tension forces.
Called step rings, they provided the additional weight
required to redirect the horizontal forces down into
the walls below. The upper portion of the dome is
approximately 5 feet thick, but with the addition of
the seven step rings, the dome widens to more than 16
feet.

133
 ^

cross section of
Perhaps the most remarkable feature of the dome is
the Pantheon
the occulus. Just where the compressive forces are the
greatest, the architects chose not to place any
material at all. Instead, the forces are resisted by a
41/2 -foot- thick ring of bricks, called a compression

ring. Like the cross section of a round tunnel, this

ring withstands the compressive force from every
angle, allowing the enclosed space to remain open.
Instead of trains or cars, however, this opening admits
and occasionally even a
light, little rain.

The shape inside the dome a perfect hemisphere

is

with a diameter of 143 feet. This matches exactly the

distance from the occulus to the center of the floor
below. Add to this geometric perfection the daily
movement of a circle of sunlight around the interior,
as well as the nighttime view of the stars, and the
Pantheon goes well beyond being just a temple to the
gods. a representation of the heavens themselves,
It is

firmly fixed at the center of Hadrian's empire.

135
 Hagia Sophia
Constantinople (Istanbul, Turkey), 532 — 537:
Once Like the rest of the building, the dome was built of
again the leader of the now Christian empire was in 2 inches thick and approximately 25 inches
flat bricks,

need of a little public relations help. Having put down square. They were laid with very thick mortar joints
riots that left 35,000 people dead, the emperor over a wooden centering.
Justinian, following Hadrian's example, turned his To admit more light into the space, the architects
attention to the creation of a great domed set windows into the base of their dome. This meant
building — in this case a church, to be called Hagia that they could not use continuous step rings, like
Sophia, or Divine Wisdom. those of the Pantheon, to counteract the horizontal
The Anthemius of Tralles and Isidorus
architects, thrust. Instead, they buttressed the brickwork
of Miletus, began by dividing a 200-foot square into between the windows with segments of step rings and
three parallel rectangles. This created a wide central strengthened the dome with shallow ribs. Without the
space for processions and two narrower aisles for benefit of continuous step rings, the remaining
viewing. In the middle of the wide rectangle they tension forces from the dome, and from the four
marked off a square a little over 100 feet along each arches supporting it, had to be restrained in some
side. Above this square they located the dome. other way.
Supporting a round dome above a square created Along the main axis of the church, the architects
challenges that Hadrian and his architects had been placed two half domes, one against each of the main
spared with their cylindrical wall. Anthemius and arches. These in turn were buttressed with a sequence
Isidorus first had four massive limestone piers built, of smaller semi-domes and vaults. Working together,
one at each corner of the central square. From the this system of masonry shapes carried all the
tops of these piers they built four enormous brick principal forces down through the piers and walls to
arches, one over each side of the square. Above each the foundations. To resist the forces pushing perpen-
pier, the space between the arches was then filled in dicular to the main axis, the architects built four
with masonry, creating a slightly spherical triangular massive rectangular blocks right up against the main
shape called a pendentive. When finished, the tops of piers. They then filled the spaces below the two main

the four pendentives formed a continuous circular arches with a second smaller arch and a combination
base for the dome. of walls pierced with windows and colonnades.

dome of Hagia Sophia

brick dome
lead covering

segments
step rings
137
 Just over twenty years after Hagia Sophia was
completed, earthquakes brought down part of the
dome and one of the semi-domes. When Justinian
ordered the church rebuilt, his new architect recog-
nized that one of the main problems had been the

cross section along main axis

tremendous tension forces created by the shallowness
of the original dome. He reduced those forces by using
a more hemispherical shape, and that is what we see
today.

139
 Sehzade Mosque
Istanbul, Turkey, 1544-1548: When Constantinople By this time, the general plan of amosque was well
was invaded around 1450, Hagia Sophia suddenly established. It included an exterior courtyard, a high
went from being a Christian church to being an central prayer chamber, and one or more minarets.
Islamic mosque. And a hundred years later, when the With the plan more or less a given, Sinan concen-
great Ottoman engineer and architect Sinan under- trated on enclosing the main interior space with
took the design of several new imperial mosques, he efficiency and elegance. At Hagia Sophia, all the forces
undoubtedly looked to Justinian's one-thousand-year- from the great dome and its supporting arches were
old creation for inspiration — not only for its size and held in check by two different systems, one along each
forms but for its durability. axis. At Sehzade, the same system of smaller domes
and semi-domes was used along both axes, creating a
symmetrical structure. The external buttress piers
were also smaller and more discreetly integrated into
the building.

central prayer chamber

exterior
courtyard

v=^^ M/ffii^ rS0 [^ Eodll E5q

=^ ^^ iJ Li
 —
St. Peter's basilica
Vatican City, Italy, 1585-1590: Since 1506, just about Unlike the dome of the Sehzade, which seems to
every Italian architect and artist of note had been grow naturally out of the structure beneath it, the
called into service to help in the building of the great dome of St. Peter's appears to be a very independent
basilica of St. Peter's. Having already contributed his structure for which the building below is merely a
Sistine Chapel ceiling, Michelangelo returned to starting point. It is divided into several distinct
Rome in 1546 to take over as chief architect. He was layers. — —
The lowest one the base rests on the main
now seventy-two. The massive central piers of the pendentives.The next two layers, the first of which is
basilica's crossing were already standing and the ringed with double columns, make up the drum.
exterior wall was under construction. His main hope Springing from the drum is the cupola, which in turn
was to unify the entire structure beneath a supports the lantern.
monumental dome. Perhaps in admiration of Hadrian's dome,
Michelangelo designed his to be 6 feet less in
diameter. But what he gave up in breadth, he more
than made up for in height. The base alone began 10
feetabove the height of the entire Pantheon, and the
complete distance from floor to lantern top was
approximately 450 feet.

lantern

cupola

drum

\\ \
?.' ^. ..==, '.^c .
N^.'-^
c^-iiTtmid , \Ul \'

141
 Also influencing Michelangelo was a remarkable Michelangelo's death. Although the elliptical shape of
dome and a half earlier by fellow
built only a century Brunelleschi's dome created less outward thrust than
Florentine, Filippo Brunelleschi. To cover the large the hemispherical shape of St. Peter's, both were
crossing of that city's cathedral, Brunelleschi had ringed by a number of iron chains for additional
constructed an octagonal dome with a thick inner stability.

shell and a thin protective outer shell. Both shells One of the most extraordinary things about the
were made of brick and were tied together by a stone Duomo is that it was built without scaffolding.
grid of vertical ribs and horizontal bands. This Brunelleschi accomplished this by tying the masonry
brilliant design not only reduced the weight of the together in such a waywould act as a compres-
that it

structure,it made it stronger and easier to maintain. sion ring even though the uppermost course of bricks
So famous was Brunelleschi's solution that the entire was still unfinished. However, the more independent
building is referred to simply as the Duomo. inner and outer domes of St. Peter's cupola were
Both shells of the Duomo follow thesame elliptical probably built using some kind of temporary
contour. Only the outer dome of Michelangelo's centering. This would most likely have been
cupola is elliptical. The inner dome is hemispher- supported from the drum rather than the floor. In any
ical —more like the may be
Pantheon. This difference event, Michelangelo didn't bother himself with such
due to Giacomo della Porta, the architect who actually details, relying instead on the ingenuity of the
oversaw construction some twenty years after craftsmen who would construct his masterpiece.

,Q

i I

brick pattern for self supporting

sides of dome

cross section of Brunelleschi's dome in Florence

143
 Les Invalides & St. Paul's
Paris, France,1680-1691: While the dome of St.
Peter's would remain the basic model for years to
come, the continuing desire to build domes high led to
a variety of inventive modifications on the inside.
When J. H. Mansart built his dome over the church of
Les Invalides, he decided on a three-part structure.
The lowest one, a true dome with an occulus and a
is made of stone. Above it rises a
hint of coffering,
second stone dome, the great weight of which is
seemingly reduced, if not eliminated, by an assort-
ment of painted clouds and deities. Mansart increased
theamount of stone around the bases of both domes
to counteract any outward thrust. Rising from and
—
protecting the upper dome is the cupola in this case
a dome-shaped timber roof sheathed in lead.

London, England, 1675-1710: In Sir Christopher

Wren's remarkable construction over the crossing of
>A/AJ>^ St. Paul's Cathedral, an inner dome of brick, once
?fpMS^pffe again complete with occulus, stands some 214 feet
above the ground. Between it and the outer cupola is
a towering brick cone. Wren obviously chose this very
strong shape primarily to hold up the massive stone
lantern, which is estimated to weigh more than 800
tons. The cone is ringed with four iron chains to resist
any tension created by its own weight as well as that
of the lantern pushing down on it. The combined
weight of the cone and lantern, along with that of the
upper portion of the drum, counteracts the horizontal
forces around the sides of the real dome.

fe±
[4
 The U.S. Capitol
Washington, D.C., 1856-1863: Near the turn of the
nineteenth century, the symboHc, or perhaps propa-
gandistic, possibihties of the dome were recognized by
none other than George Washington himself.
Although he'd never actually seen a dome in person,
he believed that a new country's most important
building— its capitol— would be taken a lot more
seriously with a dome on top.
The dome, completed in 1824, combined
first capitol

an inner masonry shell modeled on the Pantheon with

a dome-shaped roof of timber and copper cladding. At
the urging of President James Monroe, this outer
structure had been builtmuch higher than it had
been designed to be, creating an architectural
compromise that never completely satisfied anyone.
When a devastating fire in the Library of Congress
almost reached this vulnerable and not terribly
popular roof, the decision was made to give the
central rotunda a completely new covering.

protective
outer roof
original
dome

f^/^

r 1
145
 The first drawings of the new dome were created by Meigs had a better idea. He had the upper portion of
an architect named Thomas U. Walter. While the rotunda wall rebuilt and reinforced without
traveling in Europe, Walter had visited St. Peter's, alteringits diameter. On top of the wall, extending

Les Invalides, and St. Paul's, among others, and his almost 10 feet from the outer edge, he placed a ring of
design was certainly inspired by them. But it was also large iron brackets —
two for each of the thirty-six
influenced by St. Isaac's Cathedral in St. Petersburg, columns that would surround the drum. The brackets
Russia —a building Walterhad only studied in were tied together by a thick iron ring and buried in
drawings. All four structures employed the traditional the masonry. To create the appearance of a sturdy
base, two-part drum, cupola, and lantern, but St. stone base, he then had a thin curtain of cast iron
Isaac's had the additional distinction of being made of installed between the ends of the brackets and the
cast iron. Cast iron could create the look of the gi'eat roof below.
domes of the past, and it could do it far more econom- Since the entire structure was being bolted together
ically than masonry. It was the ideal solution for an piece by piece, no traditional centering was needed.
impatient young country wishing to import a little Only a single timber tower was built from the floor of
traditional respectability —and
was fireproof
it the rotunda to support a boom and the necessary
Under the watchful eyes of Walter and Captain tackle for lifting.
Montgomery C. Meigs of the Army Corps of Moving around the wall, workers guided each of the
Engineers, work began in 1856 with the removal of 27-foot-long cast-iron columns into place and secured I
the old dome and the installation of a temporary roof them to the base.About 6 feet behind each column
Because of its increased diameter, the new dome and resting directly on the wall, they then erected
—
would require a new larger base the only part of the four cross-shaped cast-iron columns. Not only would
structure Walter intended to build of stone. Because these form the base of the ribs, which in turn would
of its dimensions and tremendous weight, however, support the dome, they would also anchor the wall
this would have meant making major and costly that enclosed the drum.
changes to the existing structure, right down to the
foundations.

cross section of top of

rotunda wall
147
 temporary roof
, (not shown in
^ \A large drawings)
V, \
^ s \

Once the lower level of the drum was finished,

workers began extending the thirty-six curving ribs,
adding the ornate pieces of the upper drum as they
went. With the completion of the entire drum, the
wooden tower had to be extended so that the highest
pieces of the structure could be assembled.
In the middle of 1861, full-scale civil war broke out.
Suddenly the resources allocated for building a dome
were more urgently needed for equipping and feeding
soldiers. All contracts were suspended, and no new
materials were delivered. For the next year, the few
workers who remained assembled only those pieces of
cast iron already on hand.
Once again, though, the importance of a dome as a
public relations symbol would prevail. In 1862, as the
war raged on, it was decided that work on the Capitol
should resume as a display of determination and of
faith in the survival of a united land.
149
 Soon new pieces of the structure were arriving, and
they were quickly fastened into place. As the ribs rose,
the builders tied them together with horizontal belts
of cast iron and further secured them with cables,
whose tension could be precisely adjusted. Projecting
from the outer face of each rib were the supports to
which the pieces of the cupola were attached.
At around 200 feet above the rotunda floor, the tops
of the thirty-six ribs were connected into groups of
three. Just twelve ribs would continue on to support
the lantern. In early December 1863, the great bronze
Statue of Freedom was secured to the top of the
lantern and the exterior of the dome was finished.

151
 Work could now safely begin on the interior. An
ornate inner dome, complete with coffers and occulus,
was hung piece by piece from the ribs. A dome-shaped
canopy was then suspended immediately above it,
upon which a monumental painting would be created
to provide an appropriately dramatic vista through
the occulus. The artist, Constantino Brumidi, was still
at work on his masterpiece in the spring of 1865,
when the body of Abraham Lincoln was laid in state
in the rotunda below.

canopy for
painting
153
 Astrodome
Houston, Texas, 1962-1965: Houston needed a The dome he eventually got, called a lamella roof,
stadium large enough to house a baseball field and was built of prefabricated steel trusses fastened
seating for 50,000 fans that could be sealed off from together to form arches, which were then linked by a
heat, humidity, and insects and then air-conditioned. lattice of interlocking diagonals. Unlike any of the
Given the more-or-less circular arrangement of domes we've seen so far, the Astrodome was built
seating around a baseball field, and the need to elimi- from the center outward, using thirty-seven tempo-
nate any columns that would interfere with both rary towers. To overcome the tremendous outward
playing and watching the game, a dome was a very
logical choice. And since domes had a solid two-
thousand-year track record when it came to raising
the profiles of both places and individuals, why
wouldn't one with an unencumbered span of 642 feet
do the same for Houston? This apparently was the
thinking of one Judge Roy Hofheinz — showman,
baseball fan, and Texan.

tension ring
 thrust created in this relatively shallow structure, the
sides of the dome were contained within a strong steel
truss tension ring. Seventy-two columns below the
tension ring carry the weight of the dome to the
foundations. The actual connections between the
lamella roof and the tension ring were hinged to allow
the dome to expand and contract.
When the entire steel structure was finished
workers carefully lowered the hydraulic supports on
the tops of all the towers at the same time, and only a
sixteenth of an inch at a time, so that the dome would
settle the few inches to its final height.

one-twelfth of
lamella roof

155
^^^.- ^
 The surface of the finished dome was then covered
with acryHc skyUghts and as at any self-respecting
ballpark, theground was planted with grass. It was
not until this "eighth wonder of the world" was
actually operating that a few problems arose. The
skylights not only admitted natural light, they inten-
sified it, making almost impossible to see fly balls.
it

When they were painted over, the grass, which had

been specially developed for the Astrodome but still
depended on sunlight, promptly died. This led to the
invention of a plastic grass substitute called
AstroTurf, which was glued in thin sheets to the
unforgiving concrete slab below.
The Astrodome continued games
to offer baseball
right through the 1999 season.Now, however, the old
building has been abandoned by the Astros and their
fans for a brand-new stadium with a retractable roof
right in the heart of the city. In theory, at least, Enron
Field, as it is called, should provide the best of both
worlds.
So what do you do with an often empty though still
impressive domed stadium surrounded by a huge
parking lot with easy access from the highway? I
would use it to house the world's first international
dome exposition. What could possibly be more fun
than comparing domes side by side, and all in air-
conditioned splendor? Of course, the success of such
an undertaking would depend entirely on the
generosity of the lenders, as it would be astronomi-
cally expensive to put on —but hey, this is Texas.
Right, Judge?

157
 • I

» t

J^' C'Xr<

•::ifft:4z::;

a.j_^:_
f.-'-Ift*fr^L.-^
-r
I

m*i^-'

<-•;• ,
SISSi.'/ "

:PP ^^Wl^^ -•
?•

_^ f f t'
^'

= :x 1 -

159
 m

;o
 Skyscrapers
Over the centuries, the arduous and often fatal task

of building towers has been undertaken for many

reasons. The one shared element which often inspires
their creation is the sense of competition among their
builders. Nowhere is this more evident than in the
evolution of the skyscraper.
It all began in Chicago at the end of the nineteenth
century. In the years following the devastating fire of
1871, the city's prosperity steadily grew along with its

population, creating, among other things, a booming

real estate market. Some of the most sought-after
properties were located in the heart of the city — an
- ! area of just a few blocks bound on the north and
west by the Chicago River and on the east by Lake
Michigan. By the 1880s, the amount of remaining
downtown space could not demand. For
satisfy the
the developer or businessman, there was no longer
any alternative but to pile floor upon floor, as high
and as quickly as possible.
In 1893, when a number of buildings approached
and then exceeded the 200-foot mark, city fathers
grew nervous. Fearing that these tall, out-of-scale
boxes would turn the sunny streets into dark, inhos-
pitable canyons, they imposed a ten-story limit on all
new construction. Although this proved a significant
hindrance to the appearance of real skyscrapers in
the very city that had given birth to this quintessen-
tially American architecture (a fact for which Chicago
is still overcompensating), the high-rise office build-
ing was alive and well and ready to travel. It quickly
went east. In less than twenty-five years, the tallest
buildings in the world rose above the streets of New
York City. It would be sixty years before Chicago could
make that claim.

161
 Reliance Building
Chicago, Illinois, 1892-1895: William Hale's new
building was intended to house a number of different
businesses. The basement and ground floor would be
used as a department store. Smaller spaces on the
second floor would be offered to jewelers, tailors, and
hat makers. The upper floors were to provide the
most up-to-date offices for the gi'owing number of
doctors and dentists and their patients.
The area the building would cover 56 by 85—
feet —
was fixed on two sides by State and Washington
Streets and on the other two by an existing L-shaped
building. Because the building was started before the
new height limits were imposed, it was designed to be
fifteen stories tall. Anything taller would have
concentrated too much weight on the layer of clay
that began just a few feet below ground level.

..•Cn
 Up to this time, almost all tall buildings were
enclosed by masonry load-bearing walls. To support a
building's great weight, these walls had to be thick,
particularly at the base. This imposed serious limita-
tions on the number and sizes of windows. But an
alternative was emerging in Chicago that would
eliminate the need for exterior weight-bearing walls
altogether: a three-dimensional grid of beams and
columns capable of supporting all the loads to which
a building might be subjected. These included the
vertical forces created by the weight of the floors and
their occupants, as well as horizontal forces caused
either by the wind or, in some regions, by earth-
quakes.
Hale's Reliance Building, designed by the archi-
tects John Root and Charles Atwood, was to follow
this concept. Since all the structural requirements
would be handled by a steel skeleton, the exterior
covering could be quite thin. Its only function now
was to let in light and keep the weather out.

weight of
building
and
contents
(dead
load)

earthquake i

(live load)
steel skeleton and curtain wall

163
 column
cast-iron plate

— grillage
beams

pread footing

By the time the various loads on a building reach In 1893, when the last of the tenants were finally
the bottom of each supporting column, they have gone, the old masonry building was demolished and
become a strong vertical force concentrated in a very construction of the steel frame began. As the two-
small area. Foundations have to distribute this force story sections of column arrived at the site, workers
over a wider area or, if that is not enough, to carry it hoisted them into place and riveted them together. To
all the way down to firm soil or bedrock. By choosing further stiffen the skeleton against the horizontal
the appropriate foundations, engineers hope to forces of the wind, they bolted deep beams called
minimize the distance a building will settle and to girders into place between the exterior columns at
ensure that it does so uniformly. each floor level. The entire skeleton was assembled in
Before the foundations for the Reliance Building only four weeks.
could be built, the site had to be excavated to a depth The architects designed the outer skin (eventually
of 14 feet to expose the firm clay upon which
would it to be called a curtain wall) as a series of horizontal
rest. However, the process was made complicated by bands which alternated between high windows and
the fact that there was already a building on the narrow molded pieces of terra cotta. These glazed clay
site —
complete with tenants. But Hale was nothing if blocks not only gave the building's fagade some
not persistent. Rather than wait for the occupants' decoration, they also protected the skeleton from fire.
I
leases to expire, he decided to support the occupied Although steel will not burn, it will weaken if exposed
floors on temporary beams and columns so that his for too long to very high temperatures. Such improve-
workers could safely remove the ground from beneath ments in fireproofing, along with the provision of
them. adequate escape routes and a reliable water source on
Since the columns of the building could not rest each floor, represented another important innovation
directly on the clay, each had to have its own in skyscraper development.
especially constructed support. Called spread foot- One more thing helped nurture this new architec-
ings, these p3rramid-shaped structures were built up tural —
form the development of safe, efficient eleva-
in layers, starting with a thick concrete pad poured tors. No one understood better than Hale that stairs
right on the clay. Two or more layers of steel rails or lose their charm for most people after about five
beams, set perpendicular to each other to form what stories. In the early 1880s, he and his brothers had
is called a grillage, were then placed on top of the pad bought out the company founded by Elisha Otis, the
and buried in more concrete. A thick cast-iron plate inventor of the elevator. The Hale Elevator Company
on top of the grillage carried the actual base of the claimed to offer the most advanced elevators in
column. Chicago, and the architects were only too happy to
install four of them in the Reliance Building.
165
 WooLwoRTH Building
A healthy business dimate and lots of reliable
bedrock just below the surface set the stage for New
York City's incredible foray into the world of
skyscrapers. When Frank W
Woolworth asked the
architect Cass Gilbert to design a new skyscraper, he
requested something gothic and at least 50 feet
higher than the nearest competition— the 700-foot-
tall Metropolitan Life Insurance Tower.
What he was an extraordinary building. Its
got
overall shape was to a great extent the result of the
amount of floor space that had to be fitted onto the
site. But Gilbert greatly emphasized its verticality
by
setting the bands of ornate terra cotta once again —
—
the material of choice on end and slightly recessing
the columns of windows between them.
Although built to provide first-rate office space,
with the fastest elevators and latest safety features,
the Woolworth Building was first and foremost about
height. Both its existence and its striking appearance
owed a great Frank
deal to W
Woolworth's determi-
nation to nickel and dime his way into the record
book— which he did. His building was the world's
tallest from 1913 to 1930. Beneath the skin of this
truly "modern" skyscraper, a heavily braced steel
frame rises some sixty stories on the backs of caissons
sunk deep in the waterlogged soil below.

O
o
i>

o
a
as
u
cc

h-l

o
O
 Chrysler Building
When the title of "tallest" eventually changed
hands, it left the neighborhood but not the island.
William Van Alen's Chrysler Building almost 300 —
feet higher than the Woolworth Building was, like —
its predecessor, a towering tribute to both ego and

advertising. And it too rises on a heavy steel frame

strengthened against the lateral forces of the wind by
diagonal bracing sandwiched between the elevator
shafts.
A modest skin of white and gray glazed brick covers
most of the building. The only hints that something
slightly unusual may be afoot are the oversized hood
ornaments that occasionally slow our upward gaze.
But waiting for us at the top are crescents of radi-
ating sunbursts sheathed in chrome-nickel-steel and
punctuated by triangular windows. Once our eyes
have adjusted to this gleaming madness, they arrive
at the spire whose surprising last-minute emergence
through the top of the roof guaranteed the Chrysler
Building's height superiority over its closest rival, the
Bank of Manhattan. But even this trickery was not
enough to fend off the next challenge.

167
 Empire State Building
New York City, 1929-1931: Having determined how of any height as long as the square footage of any
much office space the Empire State Building should individual floor didn't exceed 25 percent of the area of
provide, an approximate budget for the enterprise, the site.

and the fact that it should be the tallest building in Bound by these requirements, Lamb had to deter-
the world, the project backers turned the job over to mine how to give his clients the floor space and
the architectural firm of Shreve, Lamb, and Harmon. building height they required and at the same time
As chief was William Lamb's job to
designer, it satisfy his own wish that no office worker would be
oversee the process that would eventually give the any farther than 28 feet from a window. With each
building its final form. significant increase in floor space, he had to provide
Many factors had to be taken into consideration, more elevators, so that passengers wouldn't be left
including a few, such as city zoning regulations and waiting. The more
elevators he provided, the more
site limitations, overwhich the architects had no floor space theyconsumed. The more floor space they
control. One such regulation, intended to maintain at consumed, the more floors he had to add. The form
least a minimum amount of light and air at street eventually created by the banks of elevators grew in
level, demanded that tall buildings be stepped back at response to the building's height and regulated
certain predetermined heights based on the width of massing. By the time the calculations were done,
the adjacent streets. Another stated that the portion everyone's practical needs were realized in a building
of a building rising above the thirtieth floor could be of eighty-five stories.

/
-^ -^ >' /,
169
 Why an /?
When a beam or column bends, one side goes into in the shape of an I. The two parallel parts, called the
compression while the opposite side goes into tension. flanges, do most of the work. The piece that holds
The area along the centerline between the two sides them together, called the web, can be thinner because
experiences neither. This means that in a solid it doesn't have to carry as much force. Although the
rectangular beam there is some material right in the girders of the Reliance Building were not single rolled
middle of the beam's depth that isn't doing much. In pieces like those of the Empire State Building they —
an iron or steel beam, it is possible to move much of were built up of angles and thick steel plates riveted
that underutilized material to where it will be more together —their overall shape remains the same.
effective in resisting these forces. The result is a beam

,— neutral area
neutral area -t

I— flange

As soon as the architects had determined the workers had made sure that the columns were
number of floorsand their exact square footage, the perfectly vertical, they riveted the various connec-
engineers located all the beams and columns and tions permanently.
calculated how strong they would need to be. While After the first fewwere in place, temporary
tiers
beams can be more or less the same size regardless of lifting devices called derricks were set up. These
which floor they are on, the strength and therefore would hoist the various pieces of steel into position.
the sizes of the columns depend very much on their Every few floors, workers had to dismantle and raise
placement in the skeleton. the derricks to keep up with the building.
Once the site for the Empire State Building had As the steelworkers continued to assemble the
been cleared, it was excavated down to bedrock (only frame above them, other workers were busy making
about thirty feet below street level), and spread the concrete floors. They erected temporary wooden
footings were prepared. forms around the girders and beams and then draped
Meanwhile the various pieces of the skeleton were a heavy wire mesh over the steel to help reinforce the
being fabricated. The heaviest columns, those at the four inches of concrete that would serve as the base of
very bottom of the building, were built up from each floor.

I-beams and steel plates. They were delivered in two- No sooner had the concrete set than another group
story sections with rivet holes already drilled and of workers began installing the limestone blocks,
brackets for the girders already in place. As soon as aluminum panels, chrome-nickel-steel trim, and
thefirst columns had been set on their bases, girders windows that made up the curtain wall. To help
were inserted between them to stiffen the building maintain the very tight building schedule, all of these
and to support the smaller floor beams. When pieces were designed to be installed independently of
the others.
171
 In less than seven months, workers had reached the
deck on the eighty-sixth floor.
level of the observation
Although the structure now rose four feet above the
tip of the Chrysler Building's spire, they continued
the framing so they could build a 200-foot-tall tower,
ostensibly for docking dirigibles. When such aero-
batics turned out to be much too risky — only two
attempts were made —the idea was abandoned. Yet
even though the mooring mast failed on the trans-
portation front, it perfectly concludes the building's
journey to the sky. It also gave the Empire State
Building a solid 200-foot edge over its art deco
neighbor and a height superiority that would last for
forty years.
Lamb and his engineers knew that the real problem
with height isn't the rivalry between buildings so
much as the increasingly fierce battle all tali struc-
turesmust wage with the wind. Even a skyscraper as
heavy as the Empire State Building will sway a couple
of inches off center in strong winds.
While a building's sway rarely presents a structural
threat, it can test the stamina of its inhabitants,

especially those living or working near the top. Over

the last fifty years, only a handful of buildings have
grown taller than the Empire State, but many have
grown narrower, and all are lighter. Walls of glass now
cover steel skeletons instead of the heavy masonry
cladding that helped stabilize buildings like the
Empire State. Some of these newer structures can
sway more than two feet in each direction, which is
why engineers have had to develop additional ways of
stiffening their buildings.
From the beginning of skyscraper construction,
engineers used connections that attempted to
counteract rotation and flexibility at the ends of
beams and columns. In what are called rigid connec-
tions, both the flanges and the webs are attached.
This makes the joined beams and columns act more
like a single piece. On the Empire State Building,
these rigid connections were achieved with rivets.
Today rigid connections are far stronger, because they
use a combination of high-strength steel bolts and
welding. However, because they are complicated and
time-consuming to make, they are also expensive.
 Beyond a certain height, even rigid connections
won't be strong enough on their own to adequately
reduce a building's sway. This is why the engineers of
buildings like Woolworth, Chrysler, and the Empire
State strengthened the central cores by installing
diagonally braced steel trusses between the elevator
shafts. In more recent buildings, reinforced concrete
walls running the full height of the structure have
been used to create even stronger cores. Whether
built of steel or concrete, a building's core usually
houses elevator shafts, stairwells, toilets, and other
mechanical systems.
By solving the problem of sway in tall structures
from the inside out, engineers can often make the
surrounding skeleton lighter and significantly reduce
the number of expensive connections. However, in the
tallest buildings, requirements for stiffness have
forced exterior walls to reclaim their load-bearing
responsibilities.

column

column

Ided
nnection

rigid simple reinforced concrete core

connection connection

173
 7/

John Hancock Center

Chicago, Illinois, 1969: There are two obvious things
about engineer Fazlur Khan's John Hancock Center
that remind us of the ongoing battle between all tall

structures and the wind. The first is its shape. The

second is the enormous exterior cross-bracing.
The shape both familiar and logical. At Garabit,
is

that other master engineer Gustave Eiffel made the

towers under his railway track wider at the bottom
and narrower at the top to increase their stability. Of
course, he had the additional advantage of being able
to let the wind pass through the structure. Although
Chicago, too, has something of a reputation when it
comes to wind, this approach would not have been
received too well by the residents of this 1100-foot-
high building.
The cross-bracing around the Hancock Center is

also very logical, though perhaps less familiar at this

scale. Like any good engineer, Khan was always
looking for ways to reduce the amount of material his
buildings would need. By tying continuous columns
and occasional floor beams together with enormous
diagonal wind braces, he significantly increased the
building's rigidity. And by transferring more of these
loads to the outside, he was able to reduce the
number of columns required inside.

jL:
 World Trade Center
New York City, 1972: Architect Minoru Yamasaki and The exterior wall was built in sections, either 24 or
engineers John SkilHng and Leshe R. Robertson took 36 feet high (two or three stories) and three columns
on vertical weight and horizontal forces in a different wide. As in the Reliance Building, they were con-
way. In their 1360-foot World Trade Center Towers, nected at staggered heights so all the joints wouldn't
the load-bearing exterior columns have become the be on the same level, which could slightly weaken the
wall. They are spaced only three feet apart and are wall. Each column was eventually sheathed on the
tied together at every floor by a deep horizontal beam. outside with an aluminum skin, into which a stainless
The result is a strong lattice that completely encloses steel track was recessed to guide the window-washing
both towers with a very stiff square tube. The core platform.
is also a very strong tube. Floors connect the inner The floors also arrived in prefabricated sections,
and outer tubes and bridge the distance between some as large as 60 by 13 feet. They were covered
them, creating large office spaces completely free of with a light steel decking over which the concrete
columns. floorwould eventually be poured. All the pieces of the
structure were lifted into place by one of four
climbing cranes housed in elevator shafts in the
central core. Each time the building reached the level
of the cranes, they were raised to a new height by
powerful hydraulic jacks.

North Tower

South Tower

! i ii
 Sears Tower
Chicago, Illinois, 1974: If one tube is good, then nine
must be better. The Sears Tower, also engineered
by
Fazlur Khan of Skidmore, Owings, and Merrill,
finally brought the title of world's tallest building
(1454 feet)almost literally into the back yard of the
Reliance Building. (The tower lost it in 1997 by a
mere 22 feet, but in the record business a foot is foot.)
The idea here was to use several smaller tubes and
bundle them together, creating a single huge
footprint 325 feet square. As the building rises, all but
one of the outer tubes drop off at certain heights,
creating a stepped effect that is both structurally
sound and psychologically reassuring. Only two of the
tubes, clinging to each other for dear life, rise the full
height.
Instead of spacing a lot of smaller columns close
World Trade Center, Khan enclosed
together, as at the
each tube with huge columns set 15 feet apart and
connected by 42-inch-deep girders. Deep trusses span
each tube at every story, once again creating un-
obstructed floor space. The entire building is clad in a
combination of bronze-tinted windows and black
aluminum panels, which suggest the locations of the
columns but completely obscure their scale.
For a building of this size, spread footings on clay
were not going to do the job. Each of the main
columns rests instead on a 7-foot-in-diameter con-
crete pier three levels below the street. The piers
travel 60 feet down to bedrock, where they are firmly
embedded.
177
 tuned
mass
damper

^ ^-=J-

7^
^
%i
yL
:k 7^
^ 2
^^ ¥"
#
X
^U^_
s t
71
^ H.
^^ ;^
<v ^
^^ #
^ ^
:^ ^
::ixJ^^^

:^ a2
^
^/. L

m \ \ a/
 Citigroup Center
New York City, 1977: There are a number of in- When a computer system detects an increase in the
teresting things about the former Citicorp Build- wind, oil is pumped out through the disks elevating
ing (now Citigroup Center) designed by architect the block ever so slightly. As the building moves, the
Hugh Stubbins and structural engineer William Le damper is activated by a series of pistons and starts to
Messurier. For one thing, there is the way in which follow, but because of the oil, it has difficulty catching
the top of the building has been lopped off at a forty- up. After several seconds the building reaches the end
five degree angle, givingan immediately identifi-
it of its swing and begins moving back in the opposite
able shape. Then there are the four large columns at direction. The damper will continue to travel in the
the base which have been moved from the corners, first direction for a short time until various springs
where you might expect to find them, to the middle of and arms stop its movement and send it back after
each side. This was done to open up space at street the building. Try as it will, it can never actually catch
level and create room for a new church which replaces up with its host until both have stopped moving. But
the one demolished to make room for this building. by maintaining this slightly out-of-sync choreography,
Although it is hidden behind the aluminum and glass the tuned mass damper reduces the building's sway
skin, inverted triangular bracing carries the loads on by almost half
the exterior walls to those columns. And finally there Numerous types of dampers are in use, not only in
is the way in which this building addresses the buildings but at the tops of the tallest suspension
problem of sway. bridge towers. One designed for a very tall (yet so far
The key is something called a tuned mass damper, unbuilt) skyscraper in Paris would use a 600-ton
the centerpiece of which is a block of concrete 30 feet pendulum, a portion of which would hang in a large
square, about 6 feet high, and, most important, tub of silicon. As the building sways, the pendulum
weighing 400 tons. (For obvious reasons, it was will try to keep up, only to have its movement
poured in place before the top of the building was severely restricted by the extremely dense silicon.
enclosed.) This block is supported by a three-foot- Through this deceptively simple device, the energy of
high concrete base, which sits on twelve 24-inch-in- the wind is actually being transferred to the silicon,
diameter disks. The whole contraption rests on a sparing the building some of that force and reducing
smooth concrete bed in the middle of the sixty-third the sway in the process.
floor.

main block

concrete bed

control arm
piston

simplified view of tuned mass damper

179
 Petronas Towers
Kuala Lumpur, Malaysia, 1993-1997: In 1991, a
number of world-famous architectural firms were
invited to submit designs for a pair of skyscrapers
to house the offices of the state-run Petronas Oil
Company. The two towers, which were to be of dif-
ferent heights, would serve as the centerpiece of a
new urban redevelopment project. In Cesar Pelli and
engineer Charles Thornton's winning design, the
towers, which rise like giant minarets, are both 1476
feet high. In fact, their symmetry is emphasized by
their side-by-side placement. A two-level footbridge at
the forty-first floor not only eases the traffic flow
between the towers and provides an important escape
route if one is needed, it also gives the entire struc-
ture the appearance of an enormous symbolic gate-
way.
The shapes of the floors themselves evolved from an
eight-pointed star —
a familiar form in Islamic design.
However, when the core space was subtracted, partic-
ularly at the higher levels, there wasn't enough
usable floor space left. This problem was eventually
solved by cantilevering a curved projection between
each of the points of the star.
Because of their height and relative slenderness,
the towers had to be very stiff Two attached cylin-
drical structures called buttress buildings provide
some stability while ensuring that the square-footage
requirements are met. Stiffness of both skyscrapers
was primarily achieved by the use of high-strength
concrete. In all concrete buildings, the connections
between beams and columns are automatically rigid.
Each tower is basically a tube within a tube. The
outer tube is a spacious grid of thick circular columns
(8 feet indiameter at the base) tied together at each
floorby a continuous ring beam. The lateral forces
working against the building are carried between the
buttress outer tube and the thick walls of the inner core by the
buildings
steel-framed concrete floors and by deep concrete
beams about halfway up called outriggers.

J
 /

cantilevered
projections

column
placement

181
 T~

The Petronas towers may be an impressive sight, but

it is unfortunate that we can't see what goes on below

them. The foundations of these two buildings are at reinforcing

dramatic as the structures they support.
least as for core
Knowing from the beginning that the soil in this
area was weak, the engineers initially planned to
support the columns of both towers on pairs of
concrete piers that would carry their weight all the
way down Further investigation revealed,
to bedrock.
however, that while the piers at one end of the founda-
tion would have to go down only 50 feet, those at the
other end would have to descend almost 600 feet.
Since piers shorten over time (as do both concrete and
steel columns), there was some uncertainty that the
resulting settling would be uniform. A leaning Italian
belltoweris one thing, but no one was willing to take

the chance with such an expensive and prestigious

—
national symbol. Fortunately, the site a former
racetrack —was large enough so that the towers could
be moved to a more favorable location.
Eventually, the engineers decided that each struc-
ture should rest on a thick reinforced concrete mat
that in turn would be supported by approximately a
hundred rectangular friction piles. These piles varied
in size,but the largest were approximately 4 by 8 feet
^^
in area and 400 feet deep. The pressure that builds up
around the as the weight of the building
piles
squeezes the between the bottom of the mat and
soil

the surface of the bedrock keeps the foundation from

sinking. The friction is so great, in fact, that while the
bottoms of the piles follow the profile of the bedrock,
they don't need to touch it. Workers injected grout to
form bumps on the sides of the piles, which further
increased the amount of friction.

;x

foundation mats for towers and buttress building$

 reinforcing
for column

reinforcing for
friction pile'

(^1 ^*' mit^tt i I -.^itJUfc

183
form for core

form for
ring beam
Concrete was chosen as the primary material for the
Petronas towers because of the tremendous strength
and rigidity it offers, and also because it is a familiar
building material in Malaysia. The core walls were
built first, using an enormous form that could be
jacked up a level after each pour The columns were
formed next, and then the ring beams that tie them
together Last, steel beams and cantilevered supports
were installed and the floors slabs were then poured
over them. To keep up with demand and ensure that
the concrete would all be of the same high quality, a
number of concrete mixing plants were built on site.
As the buildings continued to rise, workers near the
bottom were already hiding the rough concrete
surfaces behind a smooth skin of stainless steel and
tinted glass.
The masts at the top of each tower were jacked into
place from inside, much as the mast of the Chrysler
Building had been. One wonders if the architect isn't
already anticipating the arrival of the world's next
How difficult would it be, after all, to
tallest building.
sneak a few more sections up the elevator shafts? In
the meantime, for practice, the two towers can just
compete with each other.

core

columns

stairways
and elevators
 COMMERZBANK FRANKFURT
Frankfurt am Main, Germany, 1991-1997: Right Knowing the amount of square footage required, it
from the beginning, the building that would occupy was a fairly straightforward task for architect Sir
this site was intended to be open in the middle and to Norman Foster and his design team (which included
provide some kind of interaction with nature. The the engineers Arup, Krebs and Kiefer) to then
designers were required to give the bank's employees number of floors. On
estimate the their own, however,
access to natural light, fresh and a view, no matter
air, these numbers were not likely to produce a very
where their desks happened The building was
to be. distinguished skyscraper. And if nature, in the form
also supposed to be inviting from the outside, rather of a garden, remained at the bottom of the resulting
than imposing and aloof like most corporate atrium, no one above the fourth floor would really
headquarters. The actual footprint or shape of the notice it.

building on the ground could have been just about In order to bring the garden to the workers, the
anything. But in the end, a slightly rounded equilat- architects clumped the floors into blocks and then
eral triangle seemed to offer the best possibilities. It separated the blocks vertically with spaces large
would provide natural shafts at each of the three enough were two problems
for trees. Initially, there
corners for elevators, stairwells, and mechanical with this approach. First, only the lowest and highest
systems, creating an interior space or atrium free of floors of each block would have any visual connection
obstruction.

garden

floor plan

ventilation
through
atrium

continuous broken staggered

floor block floor blocks floor blocks
with these gardens. Second, placing these blocks at as much of the supporting structure as possible out
the same height on all three sides of the building from the center of the building to its perimeter to
would create an inherently weaker structure (the increase stability. They did this by placing all of the
same problem we saw avoided by the staggering of the building's columns in the three corners along with the
exterior columns at the World Trade Center). Both elevators, stairwells, and utilities. In essence they
problems were eventually solved by arranging the created three individual cores.
blocks in a kind of spiral pattern. Now everyone in the Each core contained two vertical steel trusses called
building, no matter what were on, would be
floor they megacolumns and three smaller columns of triangular
able to look across the atrium, through a garden, and cross-section. Once the columns were tied together at
out over the city. every floor level, these cores would become very
Having determined how the offices and open spaces strong. The space between each pair of cores would
should be arranged, the engineers then had to figure then be spanned by the floors. The outer edges of all
out how to hold them up. The relative smallness of the the floors were to be supported by eight-story-high
—
building's footprint roughly 156 feet along each rectilinear grids called Vierendeel trusses. The ends of
side —compared to its height, 840 feet,was going to these trusses were to be fixed to the megacolumns.
make stiffness a critical issue. They decided to move

typical corner core

megacolumn
 Located at each apex of the triangle, the foundations
for the building consist of deep concrete boxes
supported on clusters of piles. After about eleven
months, they were ready to receive the steel framing.
Climbing cranes were erected at two of the
building's corners and a third was set up alongside
one of the main walls. The first part of the framing to
be built was that of the atrium, which would always
remain slightly ahead of the rest of the building. Next
came the megacolumns. While new pieces of steel
were being added to them, the bases of these columns
were already disappearing in a thick wall of concrete.
Unlike the columns of most buildings, the vertical
steel members of the megacolumns remained the
same size from bottom was the amount of
to top. It

steel reinforcing buried in the concrete around them

that diminished as they rose. Ten months after work
began on the steel framing, workers completed the
fifty-first and final story of the main tower.

The Sears Tower hides the scale and to some extent

even the existence of its structural frame behind a
fairly uniform surface, making the building look even
larger than it already is. The cladding around
Commerzbank, on the other hand, was intended to
acknowledge and even draw attention to the struc-
tural system. The color difference between the metal
panels and the windows is subtle enough to unify the
surface and at the same time distinct enough to
remind us of what is actually doing the work. While
Commerzbank remains an imposing building, this

particular concession to those of us on the outside

does give it a slightly friendlier face.
Unhke most skyscrapers, in which each floor is
sealed off from the next and all of them are sealed off
from the outside, Commerzbank relies on its atrium

to create a dramatic spatial link between the various

levels,one that delivers natural light into the heart of
the building and promotes natural ventilation, since
the windows in each of the garden terraces can be
opened.

^8
189
 Although they are certainly technological triumphs,
of all the big things we build, skyscrapers are my least
favorite additions to the landscape. Those Chicago
officials back in 1893 were right. When these massive
structures rise side by side, they really do turn our
city streets into dark, windy, inhospitable canyons.
Even taken individually, these buildings seem intent
on overwhelming, belittling, or at the very least
ignoring us. The problem is, it is not so easy for us to
ignore them.
Understandably, the primary influence on most of
the skyscrapers discussed in this chapter has been the
need to get as much floor space as possible out of their
respective sites and still stay within the budget. This
is not to say that these concerns were not high on

Foster's agenda when he began thinking about

Commerzbank. But since his building gives me a little
hope, it seems like a positive and yet realistic way of
wrapping things up.
When additional concerns were introduced into the
basic design process, it became necessary for the
architects and engineers to avoid a purely mechanical
approach. In the end, the client and its design team
worked together to create a different kind of
skyscraper. Instead of raising yet another late-
twentieth-century medieval tower, their collaboration
suggests that it is possible to create a tall building
that stimulates and nourishes those who work inside
and welcomes rather than alienates those who live on
the outside.
It would appear that building big successfully,
especially when it comes to architecture, is more
likely to occur when imagination is not reserved solely
for the solution but is employed from the beginning in
the framing of the problem.

/-/
T ilv I

191
 Glossary
Arch—A curved structure that converts vertical force into Live load — The weight of a structure's temporary, movable
angled forces that travel down through its sides to a foundation. parts or contents, as well as the forces created by wind, rain,
and earthquakes.
Bedrock — The solid crust of the earth, which often lies several
hundred feet below the surface. Load-bearing wall —A wall designed to support all or part of
a building's weight.
Bending—A combination of forces that causes one part of a
material to be in compression and another part in tension. Pendentive —A slightly spherical triangular shape used to
form a continuous circular base dome above
Caisson —A watertight chamber in which people can work
space.
for a a flat-sided

under water.

Cantilever —A structure that projects beyond its point of

Pier—A vertical support such as a column.
support. Post and beam —A simple structure using horizontal beams
and vertical posts or columns.
Cast iron — Iron that has been melted, then poured into a form
of the desired shape and cooled. Reinforced concrete —Concrete with steel bars embedded in
improve resistance to tension.
Centering —A temporary form over which an arch or vault is
it to

built. —
Shaft A vertical passage dug to the level of a proposed tunnel
through which workers, materials, and equipment can be
Climbing crane —
A crane that can be raised or lowered to
moved.
keep up with the construction or demolition of a building.

Cofferdam —A watertight enclosure or barrier constructed in —

Slurry A watery mixture of insoluble material used in
trenches to prevent collapse by equalizing the pressure
a river. Water is pumped out of the enclosure or from behind the
between the trench and the surrounding soil.
barrier so workers can reach the riverbed.

Compression —A pressing force that squeezes material

Spoil —The material removed in the process of excavating.
together. Steel —An alloy of iron and carbon that hard and strong and
is

can be pounded or rolled into desired shape.

—
Concrete A building material made by mixing stone or sand
with cement and water. Concrete is very strong in compression —
Structural skeleton A three-dimensional grid of beams and
but very weak in tension. columns designed to carry all the loads in a building.

Curtain wall —A non-weight-bearing wall used to enclose a TBM (tunnel boring machine) —A highly automated piece
structural skeleton. of machinery that excavates a tunnel face, removes spoil, and
places the lining.
Dead load— The weight of the permanent, nonmovable parts
of a structure. Tension—A stretching force that pulls on a material.
—
Derrick A lifting device made up primarily of a boom and a Tension ring—A circle of material placed around the outside
mast. The base of the mast is fixed to the structure; and its top of a dome's base to prevent outward movement.
secured by either steel cables or steel-frame
is legs.
Transit —A surveying instrument that measures vertical and
Force—A push or pull on an object. horizontal angles.

Form work— The temporary mold into which liquid concrete is Truss —A rigid frame built up from short straight pieces that
poured to create a specific shape. are joined to form a series of triangles or other stable shapes.

Girder —A large beam, often built up from smaller pieces. Tunnel face — The portion of a tunnel that is continually
Girders usually support smaller beams. being dug away.

—
Grout A mixture of cement, aggregates, and water that can be Tunneling shield —A movable structure, usually cylindrical,
pumped or poured into cavities to increase strength, or that can built to protect people working in unstable ground as they
be used to create a watertight barrier, as in a grout curtain. excavate and line a passage.

Keystone —The central, wedge-shaped locking stone at the top Wrought iron —An iron alloy that is less brittle than cast-
of an arch. iron.

192
David Macaulay has translated
unusual concepts and complex
information for readers of all ages.

From the publication of his first work.

Cathedral, through the extraordinary
The Way Things Work and now with
Building Big, his superb design, and
magnificent illustration, have made
clear the logical process behind mans'
varied and complicated structures.
Whether chronicling the monumental
achievements of past civilizations or
satirizing modern architecture, he is

concerned above all with how buildings

are made and what their effects are on
people and their lives.

Books by David Macaulay

Building the Book Cathedral

TheNew Way Things Work
Rome Antics
Shortcut
Ship
Black and White
The Way Things Work
Why the Chicken Crossed the Road
Baaa
Mill
Unbuilding
Motel of the Mysteries
Great Moments in Architecture
Castle
Underground
Pyramid
City
Cathedral

Jacket art © 2000 by David Macaulav

^^^^mt