Skip to main content

Full text of "The Golden Gate bridge: a technical description in ordinary language"

See other formats


mm _ 


■r. mil /i'>?/ 







I I 

^97 ?#'////# ,r '> 


- v^u e - 



Digitized by the Internet Archive 
in 2012 




Ordinary Language 

The designing and building of the Golden Gate Bridge 
is one of the most spectacular jobs that the engineer 
has ever been called upon to undertake, and it is but 
nautral that the layman should take an interest in its 
construction. This interest is genuine, and more than 
just idle curiosity. Since engineering is an exacting pro- 
fession, its very nature calls for expression in terms of 
mathematics and technical phrases, and it is the pur- 
pose of this little book to interpret the engineering me- 
chanics involved in the building of this bridge into an 
easily understandable form. 

Through the courtesy of Joseph B. Strauss, Chief En- 
gineer, and Clifford E. Paine, Principal Assistant En- 
gineer, the author received valuable assistance in the 
form of engineering facts and data which is reflected 
in the accuracy of this presentation. 
The author feels deeply indebted to the officers and per- 
sonnel of the Golden Gate Bridge and Highway Dis- 
trict for the many courtesies extended him, and the 
splendid co-operation he received while gathering the 
material used in compiling this book. 
The field engineers of the John A. Roebling's Sons 
Company, contractors for the cable spinning, also ex- 
tended every courtesy possible. 

With the exception of the diagrams, the illustrations 
were executed by Mr. Chesley Bonestell. 

Copyright 1935, by 


San Francisco, 



The Declaration of Independence, that declared the people of the United 
States free and independent, was signed in the year 1776. That document 
united a group of 13 states that fringed the shores of the Atlantic Ocean. 
In that same year, Colonel Juan Bautista de Anza, a Spanish explorer, es- 
tablished a military post on the shores of the Pacific Ocean at the Golden 
Gate. Since the word ' 'presidio" means garrison, this post came to be called 
the Spanish Presidio. This same spot is still a garrison, having been set 
aside as a military reservation when the United States took possession of 
California, and it is still called the "Presidio" at San Francisco. 

In the same year, 1776, Padre Junipero Serra founded the Mission 
Dolores, one of the northernmost of the chain of missions that extended 
north from the present border of Mexico. The establishment of the Presidio 
and the Mission Dolores formed the nucleus from which sprang the Pueblo 
of Yerba Buena. This pueblo remained a sleepy Spanish settlement until 
our war with Mexico, during which the U. S. S. Portsmouth, under command 
of Commodore John Montgomery, sailed in through the Golden Gate and 
occupied this town. The same town that has since become the modern city 
of San Francisco. 


As seen from the map, the City of San Francisco is situated at the north- 
ern end of a peninsula, and even in the days of the Spanish occupation, the 
Golden Gate formed a water barrier that cut off expansion to the north, 
although in those days the word exploration was more appropriate. A sud- 
den impetus was given to the growth of the "City by the Golden Gate" by 
the discovery of gold, and the magic of this word drew people from all parts 
of the world. This inrush occurred in 1849, and the people who came here 
then were called the "Forty Niners." Transportation in that day was limited 
to sailing vessels and covered wagons, and the routing of the latter ter- 
minated on the east shore of San Francisco Bay. When the first transcon- 
tinental railroad was built in 1869, it followed the route of the covered 
wagons, and also terminated on the east bay shore. Through all the years, 
from the time that San Francisco was first founded, settled, and then built 
up, the Golden Gate has formed a water barrier to the north. The breaking 
down of this barrier will not only stimulate the growth and development 
of the country immediately to the north, but will open up a new outlet for 
the motorists of San Francisco, as well as promote new routes for commer- 
cial vehicles. Situated at the northern tip of a peninsula, all routes of travel 
that lead into San Francisco, save one, requires that a ferry crossing be 
made upon entering or leaving the city. 


The bridging of this Golden Gate has long been a visionary dream of 
those daring spirits that seek the solution of nature's handicaps. Now that 
engineering skill, and the indomitable will of the sponsors, have converted 
this dream into a reality, this great bridge, the largest single span in the 
world, stands unique in that it does not connect two cities. It will, however, 
connect a hinterland rich in resources and opportunities with a city that 
can be justly proud of its achievements. And the map here presented, show- 
ing this bridge and its environs, covers a territory that will soon be alive 
with a vast new flow of human activity, bridging the Golden Gate that once 
cut in two the heart of the State of California. 




Fortunate indeed is the traveler who passes out through the Golden Gate 
at the hour of sunset, for his ship literally sails directly into the sun, as it 
dips below the horizon. At no other spot can it more truly be said that the 
sun is the beauty and glory of the day, as this golden light floods the gate- 
way to the Pacific. This life giving gold is more precious than that which is 
dug from the earth, and a more apt title than that of Golden Gate would 
indeed be hard to find. Nor does this man-made structure detract in the 
least from the beauty of nature's handiwork, for it can be likened to a giant 
spider web thrown across this open stretch of water, just a full single mile 
in width, and this accompanying illustration scarcely does justice to the 
beauty of this great structure. 


The largest suspension span ever built by man is so called because the 
deck of the bridge is literally suspended in mid-air by vertical suspender 
ropes, and the motorist passing over the center of the bridge will be 265 
feet above the water surface. These vertical ropes in turn are suspended 
from cables passing over huge steel towers, extending skyward to a height 
of 746 feet above water, a height equivalent to that of a 65-story office 
building. That part of the bridge between the two towers is known as the 
center span, while those two parts on the shore side of the towers are known 
as side spans. As indicated in the plan, or lower part of the illustration, the 
center span is 4,200 feet long, and the side spans are each 1,125 feet in 
length. This 4,200 feet is the greatest single distance over a clear stretch of 
navigable water ever bridged by man, and the 746-foot towers from which 
this span is sprung are the tallest steel structures ever erected for this 
purpose. The tower on the right is known as the North Tower, and rests on 
rock at the water's edge. The tower on the left, or South Tower, rests on 
bed rock at a point 1,125 feet out from the south shore. The vertical clearance 
under the center of the bridge is 220 feet at high water level, an amount that 
will permit the largest ships in the world to pass under it. The roadway rests 
on steel trusses, known as stiffening trusses, which are 25 feet deep. The 
cables at the center of the bridge sag down 475 feet from the horizontal. 


This view, looking west out through the Golden Gate, shows the San 
Francisco shore on the left, and the approach that connects the city of San 
Francisco with the bridge on this south shore passes through the Presidio, 
a United States Military Reservation. The archway at the shore end of the 
side span here, bridges over the top of an old brick fort, commonly known 
as Fort Point. This fort, built in 1854, was originally named Fort Winfield 
Scott. It has served as an historical landmark at the Golden Gate for these 
many years, although it has long since become obsolete. Separated by a 
span of 83 years, it is doubtful whether a more striking contrast between 
two engineering structures can be found anywhere else in the world. 

The approach on the north side of the bridge, to the right, passes through 
Fort Baker, also a United States Military Reservation. The shore line here 
consists for the most part of sheer cliffs, that drop down into the water. 
The highway, after leaving the approach, extends on beyond this range of 
hills into the country north of the Golden Gate, and it is this country that 
the bridge was designed to serve, or bring closer to the metropolis of San 

If J 


All of man's ingenuity and resources were called upon to conquer the 
forces of nature prevailing at the two pier sites upon which the steel bridge 
towers rest. Water is not the natural element of man, and in a conquest 
with it he is decidedly at a disadvantage. The site selected for the foundation 
of this North Tower is at the base of a cliff where the tidal waters from 
the Pacific sweep in like a gigantic mill race. To gain access to this pier 
site, it was first necessary to construct a road down around an adjoining 
cliff, some 1,700 feet in length. Since the original shore line at this point 
described an irregular line across the site, most of the site was submerged 
beneath the water. In order to excavate the site, and erect the pier inside 
this excavation, it was first necessary to enclose the site with a cofferdam 
so as to keep the ocean out. 


A cofferdam is a box like enclosure, the general outline and shape of the 
one used here being illustrated in Fig. 3. The problem of erecting it was at- 
tacked on two fronts, one on the land side and the other from the sea. As 
indicated by the dotted line, the extreme left hand corner of the cofferdam 
was sunk in 32 feet of water by means of cribbing. This cribbing was made 
up by criss-crossing 8 x 12 timbers, very similar to the manner in which 
a log cabin is built up, and which is clearly shown in Fig. 1. It was con- 
structed of three, 34 foot wide, sections, and each section contained three 
tiers of cells. The cribbing itself was built on barges, upside down, towed 
to the pier site, and was turned right side up as it was dumped into the 
water at the low corner of the dam. Broken rock was then dumped into the 
cells in order to sink this cribbing to the bottom. The cofferdam, when com- 
pleted, enclosed 3 sides of an area approximately 178 by 264 feet, the fourth 
side being formed by the cliff. 

To complete the three sides of the cofferdam, rock dykes were extended 
both ways from the ends of the cribbing, the drawing in Fig. 3 being broken 
away at the center of the off-shore side to show the junction between the 
cribbing and one of the rock dykes. Fig. 2 illustrates the manner in which 
these rock dykes were constructed. Sheet steel piling was driven down ver- 
tically into the bottom, braced by timbers, and this sheet piling was then 
backed up with a rock fill. Fig. 6 shows how the edges of this sheet steel 
piling interlock, and form a wall very similar to a solid board fence. After 
the sheet piling had been driven entirely around the cofferdam, and had 
been backed up with the rock fill, a steam shovel was moved into its interior 
and the hole excavated as shown in Fig. 3. Since some leakage occurred, it 
was necessary to use pumps to dewater the cofferdam. 


Fig. 4 illustrates the manner in which concrete was chuted down an 
"elephant trunk" into the forms built on the bottom of the excavation. Con- 
crete was supplied to the chute by mixer trucks, that were driven over from 
the Marin batching plant. Fig. 5 shows the concrete block, or pier, as it ap- 
peared when completed, the steel for the tower legs being subsequently 
erected where the angle bars are shown projecting upwards. This pier 
tapers toward the top, the base measuring 80 x 160 feet, and the top area 
measures 65 x 134 feet. It is 64 feet high, extending 44 feet above the sur- 
face of the water, and 20 feet below. It contains a total of 24,000 cubic yards 
of concrete, weighing 45,000 tons, or 90,000,000 pounds. 


Most of us usually associate the word "gate" with the idea of an en- 
closure, by the closing of which the outside world can be excluded. The men 
upon whom fell the responsibility of sinking the pier for the South Tower, 
on the contrary, came to feel that the Golden Gate was a passageway 
through which a large portion of the Pacific Ocean endeavored to pass at 
every change of the tide. The world's largest land-locked harbor is some 
450 square miles in area, and four times every 24 hours the moon succeeds 
in pulling, by gravitational force, the waters of San Francisco Bay through 
this narrow gate, a little over one mile in width. The tide not only regu- 
larly ebbs and flows here, but attains a velocity of from 4 to 7% knots per 
hour, and it is only during a short 20 minute period at the turn of each of 
the four tides that the water slacks off to a comparatively quiet period. A 
great deal of underwater work by deep sea divers was required here, and it 
was only during these four short 20 minute periods every 24 hours that they 
were able to work. Even then, there was rarely ever a period when the 
waters at the surface were entirely quiet at this location. It was under such 
conditions as these that the pier for the South Tower had to be erected, and 
it was first necessary to construct a fender entirely around the pier site. 
This fender was designed to form an enclosed, ring like, seawall, the pier to 
be erected in its interior, in comparatively quiet water. Being permanent, 
the fender will protect the finished bridge pier from storms and collision 
with floating objects. 


Fig. 1 is a view looking north across the Golden Gate, showing the North 
Tower on the Marin shore nearing completion, with the hills of Marin 
County in the background. In the immediate foreground is shown the hill 
below which Fort Point is located. At the water's edge, beyond the old brick 
fort, is shown the pylon on which that part of the completed bridge forming 
the south side span will land. Just beyond this pylon, or concrete tower, is 
shown the fender in which the south pier was subsequently constructed. 
This south pier site is located 1.125 feet out from shore. 


This fender, or "giant bathtub," as it was dubbed by the workmen, is 
shown in plan in Fig. 3 ; in Fig. 2 it is shown in cross section, taken on the 
center line, or axis, of the bridge ; and, in Fig. 4 it is shown in cross section 
taken through its east and west axis. It is 155 feet wide on the axis of the 
bridge, and measures 300 feet on its east and west axis. It is shown in these 
drawings in its final form, with the details of the false work and pier struc- 
ture omitted, resting on the floor of the excavation made into bed rock, the 
bottom of which is 100 feet below the surface of the water. In the original 
plan it was intended to erect this fender without first excavating the rock. 
To accomplish this, it was planned to cut deep enough into the rock so that 
the base of the wall would rest on a series of more or less level benches, 
stepped around the outline of the fender plan to confirm to the natural slope 
of the rock. This plan called for building the fender up to a height of 15 feet 
above the surface of the water, leaving the 8 sections on the east end at a 
level 40 feet below the surface. It was then planned to float a caisson into 
the fender from this open east end, build these 8 open sections at that end 
up to the 15 foot level above water, and carry out the excavation inside of 
this caisson. A caisson is a cellular structure, usually built at some distant 

point, launched into the water, and then towed to the site of operations. Its 
function is to permit of the excavating of the waste material inside its con- 
fines while excluding that on the outside, serves as a form for the pouring 
of the concrete, is capable of being sunk downward vertically as the ex- 
cavating proceeds, and permits all of this to be carried out under the surface 
of the water. 


To carry out operations in the building of this fender it was first neces- 
sary to construct an access trestle out from shore to the site. This access 
trestle is shown in Fig. 1, leading out from the right end of Fort Point. Its 
construction was no little task in itself, as the supporting piles had to be 
driven down into solid rock by blasting, and its structure made rigid enough 
to withstand the storm conditions it was exposed to. Fig. 3 shows how a 
guide tower, or guide frame, was sunk down at the end of this trestle, and 
it was planned to erect the first section of the fender wall along the face of 
this guide frame, resting it on a bench in the natural slope of the rock. This 
trestle had been completed and the guide frame was in position in Novem- 
ber 1933, when, during a dense fog, a large freight steamer crashed into the 
trestle and carried away 300 feet of it. After this 300 feet had been replaced, 
another disaster occurred in the form of a storm that carried away 800 
feet of the trestle this time. This storm also carried away the guide frame, 
owing to the fact that both the trestle and guide frame had been weakened 
from the previous collision with the freight steamer. 


The erection of this fender and pier marks a new departure in engineer- 
ing annals, for it is the first time that an attempt has ever been made to 
erect a structure of this nature, and of such gigantic proportions, at a lo- 
cation literally out in the open sea. These two disasters, one following upon 
the heels of the other, were unfortunate, but neither of them was conclusive 
that the job could not be done. They simply delayed the work, and caused 
the engineers to change their plans to suit changed conditions. The plan of 
erecting the fender on the surface of the rock was abandoned, and revised 
plans were drawn up that called for the excavation of the entire site, in- 
cluding the fender area, prior to the placing of any concrete for either the 
pier or the fender that was to be erected inside of it. 


In such an exposed location, ordinary excavation methods could not be 
employed. The usual method is to drill holes down into the rock to be ex- 
cavated by means of power driven rock drills, charge these holes with dy- 
namite, and then blast the rock loose by detonating the dynamite. Fig. 2 
clearly indicates the exposed underwater position of the pier site, and also 
the outline of the excavation as actually made. Prior to making this excava- 
tion, the average depth of the site was 65 feet below the surface of the 
water, and the excavation was carried to a more or less level floor at a final 
depth of 100 feet below the surface of the water. Since the usual methods 
of blasting could not be employed here, one of the most ingenious methods 
ever devised for under water blasting was worked out and developed at 
the site to suit conditions there, and which will be described in the next 
succeeding plate. Some idea of the size of this undertaking may be had 
from the fact that there was a total yardage excavated of 60,000 cubic 
yards, or 1,625,000 cubic feet of rock. 



To determine the nature of the underlying rock located at the pier sites 
of the bridge, the services of an experienced geologist were engaged in or- 
der that a detailed study of the rock formation could be presented to the 
consulting board of engineers. In addition to the mass of data collected and 
presented by the geologist, a series of tests were conducted in which he and 
the engineers both participated. One of the tests made, was for the purpose 
of determining the downward pressure that this rock is capable of support- 
ing without yielding. A simple example of this sort of test is to take a block 
of wood measuring 12 inches square, long enough to weigh 50 pounds, and 
stand it on end. The square foot of surface on which this block would rest, 
would be called upon to support a weight of 50 pounds per square foot. If 
a 50 pound weight were then placed on top of the block, the surface on which 
it rested would be called on to support a downward pressure of 100 pounds 
per square foot. The square foot would be the unit area in the case of the 
block, but in the tests for the bridge, the unit area used was square inches. 


For the test, a representative section of the serpentine rock in the vicin- 
ity of the bridge site was selected, and this rock was loaded with weights, 
exerting a downward pressure of 460 pounds per square inch without yield- 
ing. The total weight that will come on the rock underlying the south pier 
was calculated to be less than 150 pounds per square inch, hence the tests, 
of which there were several, were conducted under a pressure equal to two 
and one half times as great as the pressure which will be exerted by the 
bridge on this same rock. Supplementing these tests, actual samples of rock 
near, and under, the pier site were brought up and compared with the rock 
on which the tests were made, and they proved to be identical. To obtain 
these samples, test borings were made. These borings are made with a rock 
drill, cylindrical in section, with a hollow core. The drill is rotated by suit- 
able driving machinery, and as it cuts downward into the rock, the solid 
core is preserved intact within the hollow center of the drill, and this rock 
core is brought to the surface. 


In the vicinity of the pier site, the floor of the bay is on an average about 
65 feet below the surface of the water. In order to obtain a determination 
of the surface of this bay floor, soundings were taken. These soundings were 
taken from a barge by dropping a 2,500 lb. weight, attached to a wire, down 
into the water. When the weight landed on the bottom, the wire became 
slack, and the length of wire payed out at this point measured the depth of 
the water. By plotting these soundings on a chart, an extremely accurate 
representation of the contour of the rock surface was obtained. By also 
plotting the test borings, described in the preceding paragraph, an accurate 
determination of the nature of the rock beneath its surface was likewise 
obtained. In addition to this, after the work on the south pier was well un- 
der way, a second set of soundings and test borings were made, and checked 
with the first set. All told some 600 soundings were made over an area of 
10 acres. A third series of examinations was made of this supporting rock 
under the South Pier after the fender was completed, this final examination 
being described later on. 



The story of the sailor who went rowing in the park on his day ashore is 
no more far fetched than the one about the sailor who became seasick on a 
barge, for such a case actually happened in connection with the operations 
carried out at the south pier fender site. The deck of the Ajax, the derrick 
barge used here, pitched up and down as much as 8 feet, at times, between 
the crests of the waves, and it is no secret that the deckhands did become 
seasick. Fig. 4 of this illustration shows the Ajax floating over the pier 
site, in which position it was securely anchored. A 14" diameter pipe was 
lowered vertically until it rested on the rock bottom. A frog, or 2 legged 
"tripod," was affixed to the bottom of the pipe to keep its lower end about 
10 feet clear of the rock. A timber frame was erected on the deck of the 
Ajax, shown in Fig. 1, so that as it pitched from the wave action, this frame 
would slide up and down on the pipe. 


A solid steel shaft, 8" in diameter, 20 feet long, and weighing 5,000 lbs., 
was dropped down inside of this pipe. A steel point, 2%" in diameter and 2 
ft. long, was fastened to the lower end of this shaft. As this drill point with 
the enormous weight of the shaft above it struck bottom, it started, or 
chopped a hole in the rock by impact. When successive impacts had effected 
a starting hole some 2 feet deep, the shaft was hoisted up and the steel drill 
point replaced by a chuck. Fig. 2 shows this chuck affixed to the bottom of 
the shaft, holding in its lower end a small bomb. These small bombs had a 
conical point at their lower ends, were 2%" in diameter, 2y 2 ft. long, and 
were charged with 3 lbs. of 60% dynamite. The upper ends of these small 
bombs were fitted with two lead ears, or lugs, and the bomb was set in the 
chuck by pushing it in and giving it a quarter turn, in the same manner 
that an electric light globe is inserted in an automobile headlight. The chuck 
was deep enough to provide a space above the top of the bomb, and a firing 
pin projected above the bomb. The shaft, with the bomb attached, was then 
dropped into the starting hole previously made by the drill point. The im- 
pact sheared these lead lugs off, and as the shaft came down on top of the 
bomb it drove the firing pin home. This firing pin set off a time fuse, so that 
there was an interval of about 60 seconds from the time the bomb struck 
the rock until it was detonated. In this 60 second interval the shaft was 
hoisted up until its bottom came flush with the bottom of the 14" pipe, and 
in case the bomb failed to shear off the lugs, and remained in the chuck, the 
force of the explosion did no damage to the pipe. By dropping an average 
of 15 of these bombs successively into each hole started by the steel drill 
point, a hole approximately 18 feet deep was blasted into the rock. Fig. 4 
shows the successive stages of enlarging the holes. 


A large bomb, 8" in diameter, 20 ft. long, and charged with 200 lbs. of 
60% dynamite, was then hoisted into position over the 14" pipe, as shown 
in Fig. 1, and dropped down inside of the pipe. It landed in the 18 foot hole 
previously made by the small bombs, as shown in Fig. 5, and was then driven 
deeper into the hole by dropping a steel shaft on top of it. As this shaft 
landed successive blows on top of the bomb, it was kept in line with the 
bomb by a tongue extending down into its top. Although not shown, the 
pipe was still in position during this operation. The blows delivered to the 
top of this large bomb were not of sufficient force to set off the charge of 


dynamite. After this large bomb had been driven down into the 18 foot 
hole, the end of a coil of electric wire, previously placed in its upper end, 
was brought up to the surface by a diver, and connected to a firing cable on 
board the Ajax. 


Six of these large bombs were planted, in as many holes, in this man- 
ner. The Ajax was then moved away from the site a short distance, and all 
six of these large bombs were set off. The force of these six bombs all ex- 
ploding at once broke up quite a large area of rock, which was then removed 
from the excavation by lowering dredger buckets from the Ajax. Fig. 3 shows 
the Ajax near the end of the access trestle, dumping rock debris into a 
smaller barge. In rough weather considerable difficulty was experienced in 
landing the debris in this material barge. On account of the tough charac- 
ter of this excavated rock, frequent replacement of the steel jaws of the 
4 yard clamshell buckets was found to be necessary. 


Upon completion of the excavation for the fender, a steel guide frame 
was lowered down vertically at the end of the access trestle. This frame was 
the same one that had been thrown down during the storm, and prior to its 
final setting, had been hoisted to the surface where it was repaired and 
lengthened so as to take care of the extra depth to which the excavation 
had now been carried. The function of this guide frame was to guide the 
forms into position, into which the concrete for the first section of the fen- 
der was subsequently poured. Fig. 6 illustrates the lowering of this guide 
frame at the north end of the access trestle, with the Ajax assisting. This 
steel guide frame was rectangular in shape, measuring 8 feet by 30 feet in 
its horizontal dimensions, and 125 feet in height. This drawing illustrates 
its final setting as it was used in the actual construction of the fender wall. 
The storm that occurred in November 1933 aptly demonstrated the force 
of the wave action at this south pier site when it twisted this 50 tons of 
open work steel structure off at the base and threw it down into the water. 
As it was lowered down on to the bottom, a diver steered its four legs into 
socket like pipe anchors that had previously been sunk down into the rock. 
To give it added stability, and insure it against further disaster from storm 
conditions, a concrete block was poured in and around its base, which is 
shown in Fig. 7. Its fastening to the trestle was strengthened by a heavy 
truss work, and wire rope anchor lines were run out to secure it in an easter- 
ly and westerly direction. 


Fig. 7 illustrates the manner in which the concrete forms for the first 
fender section were lowered down the face of the guide frame. Socket like 
jaws were attached to the inner face of these concrete forms, that engaged 
the vertical guide rails of the guide frame, and these rails steered the form 
to its correct position when landing on the rock at the bottom of the ex- 
cavation. A diver also assisted in this operation, and when the form had 
been landed, he adjusted the panels forming the sides of this box-like struc- 
ture so that the concrete would not leak out under the bottom when it was 
poured. The concrete enclosing the base of the guide frame served as a 
panel for the inner face of the form. While the width of the fender wall is 
27 V 2 feet, and each section is 33 feet long, these lower forms were 45 feet 
wide by 33 feet long, and the extra width provided a spread footing at the 
base of the fender wall. 



The guide frame, shown in Figs. 6 & 7, was made up and handled as a 
single unit. The steel frames used in the fender sections proper were fab- 
ricated in sections, and one section at a time lowered down. The bottom of 
the excavation as a whole was more or less level, but as the rock was broken 
up by the bombs, the surface itself was left quite rough in its final form. 
This was somewhat of an advantage, as concrete forms a better bond with 
a rough surface than it does with a perfectly smooth surface. However, as 
each steel base section of the fender was lowered into place, it had to be 
fitted to this rough surface of the rock on which it rested. Wooden panels, 
suitably weighted with concrete blocks, were then lowered down and clamp- 
ed to these frames, under water. This of course required a great deal of work 
by the deep sea divers engaged in these construction operations. 


Diving operations were conducted under the greatest of difficulties, as 
the water at the site is never still. Even in the slack periods, the flood tide 
would start running in from the south while the ebb tide was running out in 
the northern part of the channel, and thus create a strong rip tide. While 
working under water, the air hose and communication lines of the divers 
were carried several feet along in the direction of the tide, and they were in 
constant danger of having these broken by the whipping action set up. 
There was also considerable danger of being thrown against the structure. 
As a diver descends into the water, air must be pumped to the inside of his 
helmet and suit, and maintained at sufficient pressure to equalize the water 
pressure from without. This pressure under which he works is equivalent 
to % lb. per sq. in. for every foot of depth to which he descends into the 
water. The ill effects that he may suffer is not due so much to the pressure, 
but to too rapid a decrease of pressure as he comes up to the surface. To 
provide for this, he enters a decompression chamber upon leaving the water, 
and the air pressure in this chamber is gradually tapered off until it has 
been reduced to the normal atmospheric pressure of 15 lbs. per sq. in. 


The work of the divers was confined to the actual construction work 
that was necessary in building the underwater portion of the fender wall 
and pier. Any determination a diver may make of the nature of the under- 
lying rock is limited to an examination of the floor of the bay, to which he 
descends. As described in the preceding plate, the nature of the underlying 
rock was determined by soundings and test borings, carried out under the 
supervision of the geologists. The first set of test borings were carried out 
as a preliminary survey, and prior to the commencement of operations. The 
second set of soundings and test borings were taken during the time that 
operations were under way. One particular test boring was made just inside 
the area enclosed by the fender wall, and was sunk down to a depth of 159 
feet below the level of the bottom of the pier, or 259 feet below the surface 
of the water. The sample cores taken from this test boring showed a uni- 
form rock formation for the entire depth to which the drill was sunk. 
Shortly before this time, an extensive set of soundings were taken, and 
plotted on a chart. From these plottings a model was made of the area sur- 
rounding the pier site, accurately constructed to scale in every respect. In 
addition, a full report by the geologists was placed on file, covering the na- 
ture of the rock formation for the entire width of the channel. 



It sometimes happens that a beach lot is so situated that it can only be 
located at low tide, but at the location of this south pier fender site, the 
"lot" on which it is built is permanently submerged. It was on this sub- 
merged "lot" that a structure, equivalent in height to a ten story office 
building, was erected. The men shown in Fig. 1 on the working platform, at 
the end of the access trestle, are 25 feet above the surface of the water, or 
125 feet above the bottom of the excavation on which the box-like form for 
the base of the first section of the fender is resting. The placing of this form 
was described in the preceding plate, and is shown here filled with tremie 
concrete. Suspended from the stiff leg derrick at the end of the trestle is 
a tremie pipe, which carried the concrete down to the bottom of the "box." 
As the level of the concrete rose in the form, this tremie pipe was slowly 
Hoisted so that its bottom was always submerged slightly in the concrete. 
No difficulty is experienced in getting concrete to set under water when 
poured by this tremie method. The mixer truck shown at the left of the stiff 
leg derrick is one of a fleet that supplied the concrete. 


After the concrete in the base block of this first fender section had set, 
the form was stripped from it, and additional forms, 3 sections high, were 
placed on top of this base, as shown in Fig. 2. Steel frames, or bents, were 
placed inside these upper forms, and became an integral part of the section 
as the concrete was poured to additional heights. These vertical bents 
formed a rectangle in plan, and by carrying them up 25 feet above the sur- 
face, it permitted of a platform being built on top of them at the same level 
as that of the access trestle. As shown in Fig. 2, a Whirley derrick was then 
hauled out on to this platform, and was used to lower the forms at the right 
for the second 33 foot section of fender wall. Vertical guide rails, similar to 
those on the guide frame, were affixed to the end bents in each section, and 
this figure shows a form being slid down these guide rails at the west end 
of the first section. The top corners of these forms had projections that 
fitted into sockets in the bottoms of the upper forms. The forms consisted 
of steel frames into which panels were fitted, and were used over and over 
again. The forms and their panels were placed by divers. 


When the second section of the fender was completed, the Whirley 
shown in Fig. 2 was moved out on it. A second Whirley was then hauled out 
on to the first section and swung around facing east. This second Whirley 
was used to erect other sections, progressing in the opposite direction from 
the first, as shown in Fig. 3. This second Whirley is shown on the left, lower- 
ing a form, while the first has a tremie pipe suspended from its boom over 
a new section at the west end. These Whirley derricks advanced away from 
the access trestle, over which the concrete and bents were supplied to them, 
and the 33 foot sections of the fender wall were built up progressively, 
forming an elliptical ring in plan. When the two Whirleys met on the north, 
or opposite, side of the excavation, they had completed the pouring of the 
concrete up to the "minus" 40 foot level. That is, to a level 40 feet below 
the surface of the water. Seven sections in the course of erection are shown 
in this figure ; for a plan of the completed fender, refer back to Plate IV. 
The pouring of the successive sections progressed away from No. 1, and the 
final, or closing section, was No. 22. 



f I '1 fc 


Fig. 1 shows a section of the completed fender, taken on a line parallel 
with the axis of the bridge, and looking east. The top of the finished fender 
wall is 15 feet above the surface of the water, or at the "plus" 15 foot level. 
These plus or minus levels designate the distance up or down from the sur- 
face of the water. Using the heights from the bottom of the excavation, the 
first trip around of the Whirleys, described in the preceding plate, brought 
the wall up to a height of 60 feet from the bottom, or to the minus 40 foot 
level. This same procedure was carried out on the second trip around of the 
Whirleys, which brought the wall up an additional 55 feet, or to a total of 
115 feet from the bottom of the excavation to the top of the wall. This Fig. 
1 shows the fender wall completed, and a mixer truck on top of the platform 
at the left discharging concrete into a tremie pipe. Revised plans at this 
stage of the construction called for the pouring of the interior of the fender 
solid with concrete. It was poured in sections by submerging transverse 
bulkheads across the width of the fender, one of which is shown in this 


Fig. 2 shows the solid concrete brought up to a height of 65 feet from the 
bottom. When it had reached this height, the interior of the fender was 
dewatered, and the pier proper built up from this level in the "dry." After 
completion, pipes permit of the tidal flow of water to the interior of the fen- 
der, and in between it and the pier. Prior to pouring this concrete, a set of 
8 inspection wells were set in position to make the rock at the bottom of the 
excavation accessible for inspection. These inspection wells consisted of 
steel tubes, 4 feet in diameter, with a bell at the bottom, 15 feet in diameter. 
Fig. 5 shows this bell bottom with the engineers inside of it, inspecting the 
rock. Although the fender and inspection wells were dewatered, a certain 
amount of seepage occurred, and to keep the water from rising inside the 
wells while inspecting the rock, compressed air was introduced into their 
interiors. To permit entry, an air lock chamber was fitted on to the top of 
the wells, shown in section in Fig. 2. A hatch at the top was opened to per- 
mit the men to enter the chamber. This hatch was then closed, and the air 
pressure built up inside the chamber. When this pressure equaled that in 
the well proper, some 50 lbs. per sq. in., a lower hatch was opened and the 
men descended into the well on a steel ladder. This procedure was reversed 
on ascending. These wells were later filled with concrete. 


After the engineers and geologists had completed their inspection of 
the underlying rock, the forms and reinforcing steel for the pier itself were 
placed, and the concrete for the south pier poured up to a level of 44 feet 
above the surface of the water. The completed pier is shown in Fig. 4, as 
well as the details of the platform on top of the finished fender. The tracks 
on which the two Whirley derricks operated are clearly shown. After the 
forms had been stripped from the concrete, a protective sheathing of wood 
was built around it to prevent defacement of its surface during the balance of 
the erection period, and of course comes down upon completion of the 
bridge. The series of squares shown on top of the pier are steel slabs, 5 in. 
thick, on which the steel tower legs are to rest. To bring these slabs to a 
true bearing on top of the concrete, the tops of the piers were first ground 
off with a grinding wheel. Fig. 3 shows this grinding wheel in operation. 


The wheel was mounted on a carriage that was fed transversely across the 
bottom of a steel beam, the beam itself being fed in a longitudinal direction, 
thus enabling the wheel to reach all parts of the concrete to be surfaced. 
The steel slabs were set on the concrete in a mortar of red lead paste, and 
they were located in their proper positions by dowel pins, 6% in. in diame- 
ter, shown as dots in the drawing. This same procedure was carried out for 
the north pier on the Marin shore. 


It had originally been planned to make use of a caisson in sinking the 
excavation for this south pier. As explained in Plate IV, a double disaster 
caused this plan to be abandoned. However, one of the purposes of a cais- 
son is to function as a form for the concrete that is poured below the sur- 
face of the water, and it was intended to use the caisson for this purpose 
here. Prior to the completion of the fender wall as shown in Fig. 1 (Plate 
VII), provision was made for floating this caisson into the interior of the 
fender by leaving the 8 sections at the east end of the fender wall at the 
minus 40 foot level, bringing the balance of the wall up to the plus 15 foot 
level. The steel bents extending up from the wall, that supported the plat- 
form for the Whirleys, were removed from these 8 east end sections, and 
the caisson was successfully floated into the interior of the fender. How- 
ever, even with the other 3 sides of the fender enclosed, storm conditions 
were such that the caisson tossed and pitched so badly that there was im- 
minent danger that it would seriously damage the fender wall, and possibly 
destroy it before the east end could be closed in. Therefore, the caisson was 
towed out the following day and abandoned. Having played no part in the 
actual construction of this south pier, the caisson is not illustrated here. 


Upon abandonment of the caisson, the plan of pouring the concrete for 
the pier inside of the fender had to be revised to suit these new conditions. 
The steel bents were replaced on the 8 east end sections, which were then 
concreted up to the plus 15 foot level. The procedure from then on was de- 
scribed on the preceding page in connection with Fig. 1. Another intended 
purpose of the caisson was to have made use of the working chambers, in 
its interior, for the purpose of inspecting the underlying rock. If an inspec- 
tion had developed the necessity for further excavation, this could have 
been done from the interior of these chambers, inside of the caisson. Prior 
to the floating in of the caisson, a concrete mat was poured on the bottom, 
inside of the fender wall ; and the bell bottoms, shown in Fig. 5, were placed 
in position as this concrete was poured. Had the caisson been used, the in- 
spection chambers would have coincided with these bell bottoms, and they 
would have been connected together. Since the caisson was not used, the 4 
foot inspection wells above the bell bottoms were extended up to the sur- 
face of the water. 


These inspection wells, as finally placed, were so designed that had it 
been necessary to do so, excavation of the rock could have been carried out 
from their interiors. The findings of the inspection party, determined by an 
examination of the rock at the bottom of each inspection well, were care- 
fully checked with data previously obtained from the test borings. This 
survey resulted in a triple check on the bearing capacity of the rock on 
which this south pier rests, no further excavations were found to be neces- 
sary, and the inspection wells were concreted up solid. 



The building of this great Golden Gate Bridge marks off another mile- 
stone in the progress of engineering skill. By referring back to the year 
1756 we find another important milestone in the engineering profession, 
for it was in that year that the Eddystone Lighthouse was built in the Eng- 
lish Channel. It was built of huge blocks of stone, bound together by mor- 
tar. The active ingredient of mortar is lime, a substance obtained by the 
burning, in kilns, of a natural rock called limestone. The engineers on that 
project discovered that owing to the impurity of the limestone from which 
they derived the lime used in their mortar, a peculiar property was im- 
parted to it, causing it to set or harden just like stone itself, under water as 
well as in the air. Thus, a little over 180 years ago, was cement discovered 
as we know it today. From this discovery there gradually evolved the use 
of cement in mixtures of sand, gravel and crushed rock, that is known today 
as concrete. Like in all structures erected in more recent times this synthetic 
stone that we call concrete has played an important part in the erection of 
this bridge. 


In the lower reaches of San Francisco Bay there is found a deposit of 
countless billions of oyster shells that has been accumulating for centuries. 
The composition of this organic shell very closely parallels that of the min- 
eral rock from which cement is obtained, and a manufacturing plant capable 
of converting this shell into cement is located on top of this shell deposit. 
It is the product of this plant that was used in the anchorages of the Golden 
Gate Bridge. The plant is situated at the water's edge, with docking facili- 
ties. The cement itself was conveyed from this plant to the bridge site in 
bulk, through the medium of barges that were specially designed to keep 
it dry. The cement was loaded onto these barges by means of vacuum pumps, 
and towed some 25 miles to the batching plants located near the bridge 
site, where it was unloaded by another set of vacuum pumps. 


There were two of these batching plants, practically duplicates of each 
other, one being on the Marin side in a cove just inside the Golden Gate, 
and the other on the San Francisco side near the torpedo wharf in the 
Presidio. It was necessary to locate them in these sheltered positions, as the 
water at the bridge site itself was too rough to land barges alongside. The 
San Francisco plant was built on a wharf 208 feet long by 37 feet wide, on 
which bins were built for storing the aggregates. The cement was stored in 
steel silos with a capacity of 1,500 bbls. A set of batching bins and gather- 
ing hopper were built on a trestle work at the shore end, high enough to 
permit of the mixer trucks being driven underneath them. These batching 
bins were charged from the storage bins by means of conveyor belts, and 
when the proper amount of each ingredient had been weighed into the bins, 
the entire mass was released by gravity into the gathering hopper, and 
from there into the mixer truck. The truck was then driven out to the pier 
site, the meanwhile mixing the concrete by the rotation of its barrel like 
body, and upon arrival at the pier site the concrete was ready for pouring. 
The distance from the batching plant to the pier site, including the length 
of the access trestle, was about % a mile. There was a total of 78,000 cubic 
yards, or 146,000 tons of concrete used in the erection of the south pier fen- 
der, and 68,300 cubic yards, or 128,000 tons, for the pier itself. For the pier 
on the Marin side, 24,000 cubic yards, or 45,000 tons of concrete was used. 



The following is descriptive of the South Tower, which is a duplicate of 
the North Tower. As shown in Fig. 4, the base of each of the two legs of 
this steel tower consists of 97 cells, each 3% feet square and made from % 
inch thick steel plates, and each leg covers an area measuring 32 x 53 feet. 
The number of cells in each leg diminishes as the tower rises, this decrease 
in number taking place at the horizontal cross braces ; there being 53 cells 
at the 300 foot level, and 21 at the top of the tower. The base cells of these 
tower legs were set in position on the steel slabs, being straddled by the angle 
"irons" that extend down 50 feet into the concrete of the pier. Before fasten- 
ing the cells to the angles, the latter were prestressed by powerful jacks, 
and while under tension, holes were drilled through both the angles and the 
cell steel, and then riveted together. The jacks were then released, causing 
the embedded angles to exert a downward pressure on the tower legs. By 
prestressing the angles in this manner, a firm, solid bond is obtained be- 
tween the base of the tower legs and the concrete pier. 


The 21,500 tons, or 43,000,000 pounds, of steel used in each tower was 
fabricated in the east and brought to San Francisco in ships. The steel was 
made up in sections, and as it was built up the sections were fitted together 
as a cover is put on a cardboard hatbox, the overlapping edges being known 
as splice plates. In the main columns there were 404 units, averaging 45 
tons each, the largest weighing 85 tons. They were 8 x 12 feet in plan and 
45 feet tall, and were hoisted into position by means of a creeper derrick. 
This derrick is shown on the Marin Tower in Plate IV, and each of the two 
90 foot booms had a lifting capacity of 85 tons, powered by electric hoist- 
ing engines located at the base of the tower. This derrick was raised ver- 
tically, as the steel work progressed upward, by means of tackle fastened 
to the top of the finished steel and to the bottoms of the 4 corners of the 
derrick. There were 15 lifts in all, averaging 35 feet per lift, and at each 
lift the derrick was supported at the bottom by pushing 4 huge plungers 
into holes, provided in the steel, in the same manner that a bolt is shot home 
in an ordinary door lock. The various steel sections were fastened together 
in each tower with 600,000 field driven rivets. These were heated to a white 
heat in forges, and passed through pneumatic tubes to the riveting crews. 


The function of the towers is to support the cables and the weight that 
comes upon them. The cables are supported on the tower tops by means of 
saddles, Fig. 2 being a close up of one of them. A sleeve is bolted onto the 
cable where it changes from a circular to a hexagonal cross section as it 
passes into the saddle, the sleeve forming a weather tight connection at 
that point. The cover on top of the saddle forms the support for the beacon 
light. To provide for easy access to these beacon lights, and for mainte- 
nance and inspection purposes, a 3 foot square service elevator will be in- 
stalled in one of the cells in the east leg of each tower. The towers are 122 
feet wide at their bases, and the two legs of the towers are fastened together 
by cross bracing. The cross braces above the roadway are called struts, one 
of which is broken away in the drawing to show the truss work of which it 
is composed. Fig. 3 shows the interior of a cell with one of the horizontal 
diaphragms, placed at 15 foot intervals. The wire mesh screen covering these 
manholes were a measure of safety for the men using the ladders, of which 
there are 23 miles. 



In spinning bridge cables it is necessary that they be made accessible 
for their entire length. To accomplish this a footbridge is erected about 3 
feet below the position each cable will assume when it is spun. These foot- 
bridges consist of walks, supported on 1-9/16" steel wire strands, desig- 
nated as ropes here for simplicity. 


The supporting ropes for the side span footbridges were erected first, 
Fig. 1 showing the manner in which this operation was carried out on top 
of the South Tower. The ropes for the walk were pulled up from the anchor- 
age by means of a 2-wheeled trolley. The "skyline" supporting this trolley 
was first dragged out over the access trestle, its end hoisted to the tower top 
by a derrick, and fastened to a sheave on top of the saddle. The other end 
was passed over one pylon and through the other, and secured to a drum at 
the anchorage. By rotating this drum, the skyline was hoisted to the proper 
height, and formed a runway for the trolley. Hauling ropes were attached to 
this trolley in order to pull it up the skyline, bringing one of the footbridge 
ropes along with it. About 10 feet of the free end of this rope was looped at 
the trolley, and Fig. 1 shows the man in the drawing about to unfasten this 
free end and pass it through the box girder on which he is standing. This 
rope has a socket placed on its upper end, and after threading it through the 
box girder, or adjusting block, the trolley was eased off and the socket came 
to a bearing against the block. The rope was then disengaged from the 
trolley, and the latter was shuttled back to the anchorage for another rope. 

The lower ends of these footbridge ropes were fastened to another ad- 
justing block at the anchorage, which was supported on a steel bent, or 
frame, as shown in Fig. 2. These blocks were secured to the anchorage by 
2 nests of 7 ropes each, attached to eyebars embedded in the concrete. Their 
upper ends bear against the upper side of the adjusting block in a screw 
socket, and by turning these screws, the tension or sag of the footbridge 
can be adjusted. Adjusting sockets are also provided in the individual foot- 
bridge ropes to equalize the tension on them. The reels, containing the ropes, 
were mounted above this adjusting block, as the trolley pulled them up to 
the tower. 


In placing the center span footbridge ropes, they were first mounted on 
reels, which, in turn were mounted on a barge. As this barge was towed 
across the Golden Gate by tugs, as shown in Figs. 3 and 4, the rope was un- 
reeled and dropped down to the bottom of the ocean. Upon arrival at the 
south pier, hoisting ropes were dropped down from the towers, and by haul- 
ing in on these, the rope was hoisted clear of the water, as shown in Figs. 5 
and 6. The horizontal scale in these diagrams is distorted, the actual dis- 
tance across being 4 times as great as shown. The final position of the ropes 
was predetermined, and they were adjusted to a sag that brings them 
about 3 feet below the cable position. Coast Guard boats stopped all traffic 
through the Golden Gate while these ropes were being placed. These center 
span footbridge ropes were also fastened to adjusting blocks on top of the 
towers, as shown in Fig. 1, at the right. These two adjusting blocks at the 
top of the tower were fastened together by means of 2 nests of 7 ropes each, 
that were laid in curved steel channels on either side of the main saddles. 
Although not shown in this order here, the side span ropes were actually 
placed in position first. 



The description of the placing of the footbridge ropes in the preceding 
plate, and an explanation of the manner in which they were fastened at the 
top of the South Tower, also applies to the North Tower. The two foot- 
bridges, one under each cable position, are duplicates of each other, except 
for width. The west walk is 15 feet wide and is supported by 12 ropes. The 
east walk is 18 feet wide, and is supported by 13 ropes, the extra width pro- 
viding room for an escalator ; which consists of an endless moving rope to 
which the men can hold on to in walking up the steeper parts of the walk. 
To prevent the footbridges from ' 'whipping" around in mid-air, due to the 
force of the wind, they are secured from underneath by a system of storm 
cables, as shown in Fig. 2. These are marked with a series of warning lights 
for the benefit of the ships that will pass under the bridge during the erec- 
tion period. Part of the storm cable system was placed prior to the placing 
of the flooring for the footbridges, and the balance was placed after the 
flooring was in position. 


As shown in Fig. 3, the footbridge flooring is made up of sections 10 feet 
long, by bolting l 1 / 4x3%" redwood planks to light steel channels, with 
spaces left between the planks to cut down the wind resistance. The center 
of the walk, under the cable positions, was left open, and later covered with 
a heavy wire mesh screen. On the side spans, one of these sections was first 
bolted loosely to the footbridge ropes at the anchorage, and then pulled up 
a short distance. A second section was then bolted to the ropes, fastened to 
the first one, and both sections pulled up a short distance. This operation 
was repeated until there were 10 sections assembled together on the foot- 
bridge ropes. These 10 sections, called a train, were then pulled up the ropes 
to the tower top by means of a hauling rope. The men rode these trains up, 
and when the train was in its proper position, they clamped the sections to 
the ropes. This operation was carried on simultaneously on both walks at 
both anchorages. For placing the flooring in the center span, like opera- 
tions were carried out from the towers, the walk sections being first hoisted 
up to the tower tops in nests, and then slid down the ropes, as shown in Fig. 
1. Every tenth section consisted of a frame of very light steel girders, diag- 
onally braced, and the planking was only wired to them. These braced sec- 
tions served as fire breaks, so that in case of fire the planks could be rip- 
ped off and thrown overboard, leaving a gap to break the spread of the fire. 


The footbridges serve as a walkway for the men during the spinning, 
and subsequent operations, performed on the cables. They are temporary 
structures, and like a scaffolding on a building, come down upon completion 
of the bridge. A series of ropes are suspended above the footbridges, which 
support the cross members from which the hauling ropes are suspended. 
They are put up at this stage of the erection, and are more fully described 
later on. Cross walks, consisting of steel trusses, are placed at the half way 
and quarter points on the footbridges, which stiffen both walks to the ac- 
tion of the wind, as well as provide a convenient means of passing from one 
walk to the other. Vertical posts are bolted to the walks at short intervals, 
and ropes are strung along these posts to form a hand railing at the edge 
for safety ; the wiring for lighting and communications is also strung on 
these posts. To afford a better foothold on the steeper portions, every 
fourth plank projects up above the rest, and forms a cleat. 



Cable spinning appears to be a very complicated mechanical process to 
the average lay person. In reality, the principle involved is very simple, and 
this plate was drawn in such a manner as to simplify the explanation of 
this principle. The entire spinning operation consists of pulling out a con- 
tinuous length of wire in the form of loops, and the placing of these loops 
around two blocks. Fig. 1-A shows two blocks of wood, shaped like books, 
and a spool of thread fastened to the wall. The end of the thread is tacked 
to the inside of the block on the left, and a loop is pulled out to the right by 
means of a narrow spool, mounted on a spindle held in the hand. When it 
is over the block at the right, the spindle is tipped (B) , and the loop is given 
a half turn as it is placed on this block (C) . By pulling the thread towards 
the right, the upper part of the loop is stretched tight over the blocks (D) . 
By grasping the thread over the block at the left, and pulling, (E), the 
lower part of the loop is also stretched tight around the blocks, and the nar- 
row spool is then brought back to the starting position it occupied in ( A) . 
A second loop is then pulled out in the same manner, and the operation re- 
peated until the required number of loops are placed around the two blocks. 


Fig. 2 shows the set up for pulling out these loops of wire from the south 
anchorage, with the first loop being taken over the south tower by the spin- 
ning wheel. Instead of the wooden block shown in Fig. 1, the end of the wire 
is placed around a strand shoe, and this loose end clamped. The spinning 
wheel is securely fastened to the hauling rope, and moves with the rope as 
it is pulled towards the north anchorage by the hauling rope drive. The rope 
is actually supported on sheaves suspended from the hauling rope supports, 
shown in Plate XV, but only those at the top of the tower are shown here 
in order to simplify the diagram. Fig. 3 shows this same loop of wire as hav- 
ing arrived at the north anchorage, the man in the act of removing the loop 
from the spinning wheel, and about to place it around the strand shoe. In 
Fig. 1, the loop was given a half turn at the end of the trip. The wire com- 
ing from the spool is called the live wire, and this half turn places the live 
wire on top of the wheel for each trip, thus preventing the wires from cross- 
ing. On this bridge the loops were given a half turn at the beginning of the 
trip. This half turn at the beginning of the trip is easily demonstrated with 
a spool of thread, but a diagram does not demonstrate the principle so easily 
as does the one shown here. If the wheel in Fig. 3 were now returned empty 
to its starting point, it would be ready to pull out a second loop of wire from 
the same reel, and this second loop would be placed on the same strand 
shoes. Each successive trip north of this spinning wheel pulls a loop from 
the same reel of wire, and when 226 loops have been placed on this one pair 
of strand shoes, the bundle so formed is known as a Strand. Since there are 
two wires in each loop, this strand will contain 452 wires, and each cable 
is composed of 61 strands, averaging 452 wires per strand. 


To simplify the explanation of the spinning operation, only one loop of 
wire is shown in the diagrams ; actually, double spinning wheels were used, 
and two loops of wire were pulled over during a single trip of the wheel, from 
two reels. Both loops were placed on the same pair of strand shoes. Fig. 4 
shows a completed strand of 472 wires that was spun by the spinning wheel 
while making successive north bound trips, the ends near the anchorages 
being drawn to a larger scale in order to illustrate how the two sides of the 
loops were brought together after leaving the strand shoe, at a point called 
the throat. 



The first wire that is strung across from anchorage to anchorage is 
known as the guide wire. This wire serves as a guide for the balance of the 
wires, and is carefully adjusted to the proper sag. The sag is the vertical 
distance, at mid-span, from a horizontal line projected across the tower tops 
from the centers of the saddles. The amount of sag determines the amount 
of tension, or pull, in the wire, which is worked out according to a mathe- 
matical formula. To "spot" this guide wire, a transit was mounted on the 
hillside, and a line of sight taken at the point of proper sag, as shown in Fig. 
1. If the wire sags too much, the sag is reduced by pulling in on the ends of 
the wires at the anchorages ; if too little, the wire is slacked off. In order to 
afford a means of visualizing this ratio of sag to tension, without the use 
of mathematics, it is here presented in graphic form. Fig. 2-A shows a 
string stretched over the backs of two chairs to represent the cable sag. If 
the chairs are pushed apart as in (B) , the string will assume a horizontal 
position. If the chairs are pushed farther apart, the tension will become 
great enough to break the string, as shown in (C) . On the other hand, if the 
chairs are pushed together, as in (D) , the string will rest on the floor, and 
in that position there will be no tension in it at all. 


Fig. 3 illustrates how this principle is applied in adjusting the individual 
wires. In Fig. 3-A, three of the wires are shown as already having been ad- 
justed to the sag of the guide wire, the sag in the fourth wire being greatly 
exaggerated to indicate that it has not yet been adjusted. A device, called 
a "come-along,' ' is placed on this wire at a point near the arrow shown in 
(B) , pulling the slack out of the wire in the side span at the left, now indi- 
cated by a dotted line. In (C) another come-along is pulling the slack out 
of the wire in the center span, and in (D) , out of the wire in the Marin side 
span. This proceeding takes the slack out of one side of the loop, and by 
sliding the wire around the strand shoe, it is transferred to the other side 
of the loop. The other side of the loop is adjusted by pulling the wire across 
in the opposite direction, and it is then ready to again be placed on the spin- 
ning wheel. The continuity of this adjusting is illustrated in Fig. 1-D & E, 
of Plate XI. This adjusting operation follows along immediately after the 
wire has been hauled across by the spinning wheel. Men stationed along the 
footbridge at the mid-points of the spans, send an "O.K." signal to the con- 
trol station as the wire is adjusted to the correct position. 


The strand is spun from a continuous length of wire, in the form of loops, 
and is fastened at the anchorage by placing these loops around the strand 
shoes, as shown in Fig. 4. In this drawing, these shoes are 7 feet apart, actu- 
ally the two shoes are about iy 2 miles apart, with a corresponding length 
of wire between them. A single coupling makes a continuous length of wire 
of the loops shown in this figure. In the cables these couplings occur every 
3,750 feet, the lengths in which the wire is manufactured. The coupling 
consists of a sleeve, that is pressed over the two ends of the wire by a 100 ton 
hydraulic press, as shown in Fig. 5. The corrugations are pressed into the 
wire ends, shaping them to an eliptical section that prevents turning. The 
wire is coupled in this manner as it is placed on the reels, and the wire end 
of one reel is spliced to the wire end on the succeeding reel. The wire is gal- 
vanized, and is 0.196 inches in diameter. 



This plate illustrates the spinning operation as it is carried out at the 
south anchorage. The wire is pulled out over the towers by the spinning 
wheels. The spinning wheel is firmly attached to the hauling rope, and the 
hauling rope is driven by electric motors. "Shuttle Wheel" would be a more 
descriptive term, as these spinning wheels are shuttled back and forth by 
reversing the motors when the wheels arrive at their respective anchor- 
ages. The one shown here is traveling away from the anchorage, or from 
left to right in the drawing. 


The wire is delivered to the bridge site on reels, or spools, and these reels 
are placed on unreeling machines, shown at the top of Fig. 1. As the wire 
leaves the unreeling machine, it passes over a series of sheaves, mounted on 
a frame work, and one of these is known as a Floating Counterweight 
Sheave. The pull exerted on the wire by the spinning wheel varies, and if it 
becomes excessive, the floating sheave and its counterweight will move up- 
wards. If the pull eases off and the wire begins to get slack, the counter- 
weight pulls the floating sheave downwards. A man stationed on a control 
platform above the unreeling machines watches this floating sheave ; if it 
rises, he speeds up the machine ; if it falls, he slows it down. These machines 
are driven by hydraulic power and are reversible, so that if the spinning 
wheel suddenly stops, the operator can cause them to serve as a brake to 
prevent the wire from ' 'raveling". Tracing the wire vertically downward 
from the frame, it will be seen that its direction is changed by another 
sheave just in front of the "eyebars". The direction of the hauling rope is 
also changed at this point, and both rope and wire pay out on a slope that 
extends to the tower top. 


The wire passes around the spinning wheel in the form of a loop and the 
wire in the lower part of the loop then passes around the strand shoe. Fig. 2 
shows the spinning wheel to a larger scale, with a detail of the strand shoe 
directly underneath it. The strand shoe is a flat, horseshoe shaped casting, 
grooved on its outer edge to receive the wire. It is shown in section to illus- 
trate the manner in which it is pinned to the two eyebars, between which it 
is placed. The pin fits snugly in the round holes in the eyebars, but the hole 
in the strand shoe is elongated so that the shoe may be adjusted by placing 
shims behind the pin. In this figure, the strand shoe is shown in its final po- 
sition between the eyebars, the better to illustrate the manner in which the 
wire is fastened at the anchorages by being looped around these shoes ; the 
shoes in their turn are pinned to the eyebars. Actually, during the spinning 
operation, the strand shoe is placed in what is known as the spinning posi- 
tion. It is held in the spinning position by a strand leg, as shown in Fig. 3. As 
shown in Fig. 2, it would be impossible to get a loop of wire around the shoe 
while pinned in between two eyebars. This strand leg has a bracket that 
bolts on to the end of the shoe, and holds the shoe from one side only, so that 
the loops of wire may be placed in the groove of the shoe from the other 
side. After the strand of 472 wires has been completed, the strand leg is 
pulled back by means of the 150 ton hydraulic jack fastened to its other end, 
until the hole in the strand shoe matches the holes in the two eyebars. The 
shoe fits in between the two eyebars, and the pin is inserted through all 
three members. The eyebars are so-called, because they have holes in their 
ends, the same as an ordinary sewing needle has an eye in one end of it. 



There are four main saddles in this bridge, two on each tower, and each 
one is 21 ft. 7 in. long, 11 ft. high, 10 ft. wide, and weighs 150 tons. Each 
saddle was cast of steel, in three separate sections, and the ends of these 
sections were machined to a smooth surface and then bolted snugly to- 
gether. By casting them in sections they were easier to manufacture, and 
the 85 ton derricks used for the erection of the steel work were of ample 
capacity to hoist them to the tower tops, one section at a time. 


The saddles form a seat for the cables as they pass over the tower tops, 
and transfer the weight of the cables and their suspended load to the towers 
in a vertical direction. A center line at the axis of each cable, coming up 
on a slope from either side of the tower, forms an angle at the mid-position 
of the saddle, and the shape of the cable groove in the saddle forms an 
easy curve over which the cable can affect this change in direction. As 
shown in Fig. 3, the cable forms a hexagonal cross section in the saddle, 
that flattens slightly due to the enormous downward pressure at this point. 
Ribs are cast in the saddle to take care of this flattening effect. As a new 
development here, the engineers have worked out the division of the wires 
into 2 256-wire strands, 4 384-wire strands, 4 432-wire strands, 7 436- 
wire strands, 3 464-wire strands, and 41 472-wire strands, a total of 
27,572 wires, divided into 61 strands. This division of the strands into an un- 
like number of wires, together with the arrangement of placing an apex of 
the hexagon at the top and bottom, causes the cable to "lay" better. Later, 
when the cable is squeezed, this arrangement will permit of it being brought 
to a circular cross section with a minimum of vertical displacement of the 
wires from their original hexagonal formation. 


The distance between the two towers, measured along a horizontal 
line, is 4,200 feet. Since this center span is 4,200 feet long, and the two 1,125 
foot side spans, taken together, equals 2,250 feet, there will be 1,950 more 
lineal feet of roadway structure suspended from the center span than is 
suspended from the two side spans combined. This extra amount of dead 
load, i. e, steel work, roadway, etc., in the center span will cause the cables 
to sag down more in this span, and affect a pull on them towards the mid- 
span point. Since the cables are fixed at their ends to the anchorages, this 
movement towards the mid-span point will cause the sag of the cables in 
the side spans to decrease. This movement takes place as the suspended 
structure is being hung from the cables, and to maintain the towers in their 
normal vertical plane, provision is made for it by giving a "set-back" to 
the saddles when they are placed on top of the towers. This set-back is 
illustrated in Fig. 1-B and is the distance between the two center lines, one 
drawn through the saddle, and the other through the tower. Fig. 1-A shows 
how the two center lines coincide after the saddles have moved channel- 
ward. This set-back, or calculated channelward movement, is 5 ft. 6 in., 
on the south tower, and 3 ft. 7 in., on the north tower. To permit of this 
movement, the saddles are set on a nest of rollers when placing them on 
the tower tops. The movement is controlled by a set of hydraulic jacks, 
two being placed at each end of each saddle. After the bridge is completed, 
the saddles are made fast by brackets, as shown in Plate VIII, and the 
brackets against which the jacks were placed are burned off with acetylene 
torches, flush with the tower tops, as shown in Fig. 1-A. 



The spinning wheel in Plate XIII was shown as leaving the anchorage 
with its loop of wire, while being hauled up the footbridge towards the 
tower top. This plate continues that operation, and shows the same wheel, 
indicated here as No. 1, as passing over the saddle on top of the tower. The 
hauling rope is supported on a set of split sheaves that permit the spinning 
wheel arms to pass in between them. Fig. 1 shows the arrangement of these 
sheaves on top of the tower, and how they cause the hauling ropes to change 
their direction from the slope of the side span to the slope of the center span. 
Over the footbridges, the hauling ropes are supported in a similar manner, 
except that the sheave supports are hung from wire ropes, suspended from 
tower to tower, as shown in Fig. 2. 


As the spinning wheels shuttle back and forth between anchorages, 
carrying loops of wire with them, they spin four strands at one time. The 
wire is not spun directly into the main saddles, but is deposited into spin- 
ning saddles, or strand saddles, of which there are four. Assuming the wheel 
indicated as No. 1, to be north bound ; the loops of wire on it are being drawn 
out from the reels at the San Francisco anchorage, and the wire is being de- 
posited into strand saddle No. 1. These loops are carried clear on over to 
the Marin anchorage, where they are looped around a strand shoe. At the, 
same time, wheel No. 2 is pulling loops of wire out from the Marin anchor- 
age, and depositing them into strand saddle No. 4, and these will be looped 
to a strand shoe at the San Francisco anchorage. After the loops shown here 
have been pulled out to the respective anchorages, the wheels will be re- 
versed, and upon their return trip they will deposit wire in strand saddles 
Nos. 2 and 3, respectively. Strands Nos. 1 and 3 will require 118 north bound 
trips of the wheels, and strands Nos. 2 and 4, 118 south bound trips to com- 
plete four 472 wire strands ; the number of wires per strand vary, as will be 
explained later. There are 2 wires in each loop and each wheel carries over 2 
loops. The lower wires of the loops, or dead wires, do not move as they are 
laid out, and are deposited directly into the strand saddles. The upper, or 
live wires must travel twice as fast as the spinning wheel in order to lay 
out their own length, as well as a like amount of dead wire. To prevent dam- 
age to these, they are laid on a set of small wheels while running out, and 
are later lifted over into the strand saddles. 


The strand saddles, four in number, consist of a series of short blocks, 
the interiors of which are "U" shaped. Flat pieces of steel are bolted onto 
their ends, so that the wires are shaped to the circular form of the strand as 
they are deposited into them. When the strand has been completed, it is 
hoisted by means of a balance beam, and lowered down into the main saddle. 
The curve of this balance beam conforms to the curve of the saddles, and is 
lifted by two hydraulic rams, supported from above by the temporary frame 
erected above the saddles. This arrangement permits of the strand being 
lifted without bending, or otherwise damaging, the wires. For lifting, the 
strand is fastened to the balance beam by passing flat bands underneath the 
strand, and in between the blocks forming the strand saddle. Fig. 1 shows 
three of the strands in place at the bottom of the main saddle, while the 
fourth strand is suspended on the balance beam and about to be lowered 
down into position, and the spinning wheels carrying over the first loops of 
wire for the second set of 4 strands. 



Fig. 1 is a diagram, looking east, of the Marin tower, back stay, and 
anchorage. The Marin anchorage is located on the back slope of a hill that 
lay between it and the tower, and this hill was cut away in order to permit 
the roadway of the bridge to clear the ground. The engineers designate a 
cut of this nature as "daylighting." Fig. 2 illustrates, in perspective, the lay- 
out of the spinning machinery. The apparatus is practically a duplicate of 
that used at the San Francisco anchorage, except that it is arranged differ- 
ently. The difference in arrangement at the two anchorages may be seen 
by comparing this plate with Plate XIII. The hauling rope drives at the 
San Francisco end are mounted on a step, just above the eyebars, whereas 
at the Marin anchorage they are mounted on top of the anchor blocks. At 
the San Francisco anchorage the unreeling machines are mounted on the 
weight blocks, one set being on top, and the other set just below these on a 
stepped recess in the block. At the Marin anchorage, shown here, the un- 
reeling machines are located in between the two footbridges, and forward 
of the anchor blocks. The frame, containing the floating counterweight 
sheaves, is located in between the two anchor blocks here, whereas on the 
San Francisco end, these frames are mounted on the same level as the haul- 
ing rope drives. The hydraulic presses for splicing the wire is shown just 
above the unreeling machines, a like set being located at the San Francisco 


In the spinning operation, the wire is led from the unreeling machines 
up to the counterweight sheave frame ; then at right angles from this frame 
to a point on the anchor block above the eyebars ; and their direction again 
changed by a set of sheaves to bring them down to a point just above the 
strand shoes. This routing, or reeving of the wires, is shown in Plate XVII, 
and the manner in which the wire is manipulated from the strand shoes to 
the spinning wheel is illustrated in Plate XI. By means of the hauling ropes 
and drives, the loops of wire are then pulled out towards mid-span. At the 
same time, another wheel is being pulled out from the San Francisco end. 
When these two wheels, traveling in opposite directions, arrive at the mid- 
span point, they are stopped and the wires on one wheel are exchanged with 
those on the other wheel, as shown in Fig. 3. The hauling ropes are then 
reversed, and each wheel returns to its respective anchorage, but with the 
loops of wire that the other wheel pulled out to the mid-span position. 


Plate XVIII illustrates the manner in which the completed cable is 
splayed out at the splay point like the roots of a tree, and how the individual 
strands are separately fastened to the eyebars. Fig. 4, shown here, is a lay- 
out of this splay point during the spinning operation. As the spinning wheels 
pass this point, the wire is deposited in a set of strand saddles in the same 
manner as was done at the top of the tower, shown in Plate XV. When the 
strands are completed, they are hoisted by a hydraulic ram, and then low- 
ered into the splay collar. This operation being carried out at the same time 
as the strand is lowered into the saddle on top of the tower. This splay col- 
lar does not function as a saddle, by supporting the weight of the cable, 
but fixes the point at which the individual strands merge into the solid mass 
of wires that are shaped to the circular cross section of the finished cable. 
The interior of the splay collar forms a truncated cone. Upon completion of 
the spinning, the upper half is bolted to the lower half, thus clamping the 
collar firmly to the cable. (See Plate XXVIII) . 



The cable contractors have worked out a new development in wire spin- 
ning for this bridge. Although simple in principle, a detailed drawing of the 
process is extremely complicated, and it is therefore presented here in 
diagram form. While the wire is being spun over one footbridge, the fin- 
ished strands on the other walk are being adjusted and placed in the 
saddles. However, in order to show the continuity of the travel of the wheels 
and hauling ropes, the spinning is here indicated as being carried out on 
both walks at the same time. Actually, during the adjusting of the strands, 
the wheels are detached from the ropes, and the latter simply run idle over 
that walk. To switch to the spinning operation the wheels are replaced, 
and those over the other walk are detached. 


Fig. 1 is a diagram, in plan, of the general layout of the spinning ap- 
paratus. The unreeling machines, containing the wire are behind the haul- 
ing rope drives on the San Francisco side, and in front of them on the Marin 
side, and are simply designated here as wire reels. The hauling ropes turn 
behind the drives, and pass from one footbridge to the other through a set 
of frames that are equipped with counterweight sheaves, similar to those 
used on the wire. (See Plate XIII). The hauling ropes are not continuous 
from one anchorage to the other, but only run to the far tower, where they 
pass around a set of sheaves, and return to the same anchorage. This ar- 
rangement places four hauling ropes over each footbridge in the center 
span, two being driven from the Marin anchorage, and the alternate two 
from the San Francisco anchorage. The drives are reversible, and the ropes 
travel in either direction. While one part of the rope is traveling north- 
bound over one footbridge, the other part of the same rope is traveling 
southbound over the other footbridge. The spinning wheel frames are so 
fastened to the ropes, that in the position shown here, one wheel from each 
anchorage, on adjacent ropes, arrives at the mid-span point at the same 
time. As these two wheels arrive at mid-span, one wheel on each of the re- 
maining two ropes, over this same footbridge, would arrive at the Marin 
and San Francisco anchorages, respectively. No wire is shown in this Fig. 
1, and the wheels are numbered arbitrarily in order that their paths may 
the more easily be traced through the diagrams. The arrows indicate the 
direction in which the wheels were traveling just prior to coming to rest. 


In Figs. 2 and 3, only the strand shoes, wire reels, loops of wire, and 
spinning wheels are shown, and the wheels match the position they take on 
the hauling ropes in Fig. 1. In Fig. 2, wheels 1 and 2 over the west foot- 
bridge, and wheels 5 and 6 over the east footbridge, are shown as hav- 
ing pulled out loops of wire to the mid-span point. While in this position, 
the loops of wire are exchanged from one set of wheels to the other. ( See 
Plate XVI, Fig. 3) . At the same time, loops of wire are placed on wheels 3, 
4, 7 and 8. The hauling rope drives are then reversed, and the wheels hauled 
out to the position shown in Fig. 3. In this diagram, the wheels, strand 
shoes, and strands are all given the same numbers in order to trace the 
paths of the wheels, viz. : Wheel No. 1 is pulling out loops for No. 1 strand, 
that will be placed on strand shoes numbered 1. However, to indicate that 
the loops were pulled from the mid-span point to the anchorages by a set 
of wheels different from those which pulled them to the mid-span point, 
wheels 1 and 2, and 5 and 6, have been switched over from the original lanes 
in which they were shown in Fig. 2. Pulling the loops out as shown here 
would require a half turn before starting. (See Plate XI) . 



The anchorage consists of three parts ; namely, the Base Block, the An- 
chor Block, and the Weight Block. The rock upon which the anchorage rests, 
was first excavated, and a series of step like ditches were cut into the sur- 
face at the bottom of this excavation. As the concrete was poured into this 
excavation, the concrete followed the contour of these step like grooves, 
thus causing the base block to be keyed into the rock. The top of this base 
block was also stepped in this manner, and when the anchor block itself was 
poured on top of this, it in turn was keyed to the base block. The weight 
block rests on top of the anchor block, and its weight not only forms part 
of the resistance to the pull of the cables, but acts vertically downward to 
turn the pull of the cables against the rock foundation. The object of the 
anchorage is to resist the pull of the cables, which is due to their own weight 
as well as to the load that is suspended from them. The weight of both the 
cable and load acts vertically downward, due to the force of gravity. Since 
the entire mass is suspended, this vertical force is transferred to a pull in 
the cable that is parallel to its axis. Part of this suspended weight comes 
upon the towers where the cables pass over their tops, and the remainder 
is exerted as a pull at the anchorages. 


Of the several methods of fastening a rope or hawser to a stationary 
object, the most common is to bend it around a pin, one or more turns, and 
secure the loose end to the part on which the strain is placed. This also ap- 
plies to steel wire rope, except that the loose end is clamped to the other 
part. Another method for wire rope, is to place a socket on the end, and let 
this socket bear against the fixed object. Any of these methods are limited 
to ropes or cables of ordinary diameters. It is obviously impossible to bend 
a cable, 36 V2 inches in diameter, around a pin; nor is it practical to devise 
a single clamp that would hold a cable of such size. In addition to this, the 
pull on one of these cables, if concentrated at a single point, would pull loose 
from the anchorage. The cable is therefore "splayed" out at the anchorage, 
just as the roots of a tree branch out at the point where its trunk enters the 
ground. It is the individual "roots", or strands, that are actually fastened 
to the anchorage. This is done by looping the wires, that form the strands, 
around strand shoes, as shown in Fig. 4, Plate XII. 


This plate illustrates the completed San Francisco anchorage. Each 
cable has a separate anchorage, with a corridor in between the two, the near 
anchorage being cut away to show the manner in which the eyebar chains 
are embedded in the concrete, there being 3 links in each chain. The rear 
ends of the first eyebar links are pinned to girders, and the pull of the cable 
extends back to these girders, thus interposing the weight and bulk of the 
anchorage in front of this pull. The third set of eyebar links, to which the 
strand shoes are pinned, were not covered with concrete until the spinning 
was completed, and were placed, a horizontal layer at a time, as the strands 
were completed. There are 61 pairs of eyebar chains, each pair accomodat- 
ing one cable strand. Each anchorage contains 64,000 tons of concrete, or 
128,000,000 pounds, which is opposed to the 63,000,000 pound pull exerted 
by each cable. Together, these two anchorages contain 256,000,000 pounds 
of concrete. The anchorages on the Marin side are a duplicate of these 
shown here, except for the difference in the details of the pylon arrange- 




The wires that make up each strand are treated as a unit as they are 
placed on the same pair of strand shoes, and later lowered down into the 
main saddles and splay collars. After the last loop of the 61st strand is spun 
and this strand placed in its final position, the strand as a unit has served 
its purpose. The next operation is to merge these 61 units for each cable into 
a solid mass of wires. The object of squeezing the wires together, is to bring 
the cable to a compacted circular cross section. Compacted, the wires in the 
cable act as a unit, enables the cable bands to obtain a firm grip, and permits 
the wrapping to be applied to a tight smooth finish. 


As the strands themselves are formed, they are compacted into a cir- 
cular cross section of about 5 inches in diameter, and banded. The operation 
of compacting all the wires in the entire cable into a more or less solid cir- 
cular cross section, is known as the squeezing operation. This squeezing 
operation is performed by means of a powerful hydraulic jack. The jack con- 
sists of a circular yoke, or collar. This yoke is made up of two halves that 
are fitted around the cable, and then bolted together, so as to form a com- 
plete ring. A set of 12 plungers extend radially inward from this ring and 
are brought to a contact with the cable. The upper ends of these plungers ex- 
tend up into cylinders bored in the yoke. Suitable passageways are led from 
the interior of these cylinders to the face of the yoke, where they are con- 
nected to pipes, and these pipes in turn are connected to a portable hydraulic 
pump. A special liquid is used in this pump, and this liquid is forced into the 
cylinders above the plungers of the jack. As the liquid is incompressible, 
the pump is capable of building up an enormous pressure behind the plun- 
gers, which forces them against the cable wires at a pressure of about 4,000 
pounds per square inch. 


Fig. A represents a cross section of the cable, which is not unlike the 
cutting through of a loaf of bread, and looking at the end of it. This is the 
way the individual strands would appear before any squeezing pressure is 
exerted on the cable. B shows how the strands are brought together as the 
pressure is applied by means of the jack. In C the pressure has been in- 
creased, and it will be noted that the strands have begun to lose their cir- 
cular form. In a series of circles there are a certain amount of voids where 
the curve of the circles are not tangent to one another, and in this figure the 
individual wires are shown as being squashed down into these voids. D 
shows this squeezing effect as having compacted the cable to a more or less 
perfect circle, 36 V2 inches in diameter. However, just as the strands leave 
voids in between them where the circles are not tangent, so is there a cer- 
tain amount of voids in between the individual wires when squeezed to their 
ultimate density. The percentage of voids in the finished cable will average 
about 18% of its total cross sectional area, but for all practical pur- 
poses, the cable is squeezed to a solid mass. Prior to releasing the jack, a 
band is clamped around the outside of the cable just behind it, so that the 
wires will not spring out as the jack is released. The jack is started at the 
top of the towers, and is slid down on a carriage framework that rolls on 
wheels fitted to the top of the cable, the squeezing intervals averaging about 
three feet along the cable. As the cables can not be squeezed inside the 
saddles, the strands are so arranged as to effect a compact form of the 
cable inside of the cable grooves. 



The roadway of the bridge proper extends only to the ends of the side 
spans, at which points are located the two pylons, S-l and N-l, respec- 
tively. Shoreward, beyond these two pylons, the roadway structures are 
known as approaches. As the structural steel is hung from the cables, its 
weight will exert a downward pull, and as the cables sag down in the cen- 
ter span, the saddles will be pulled in that direction. At the same time, that 
part of the cables in the side spans will move towards the towers. Since 
the ends of the cables are fixed at the anchorages, and do not move, this 
pull on them in the side spans will cause them to draw up taught, their sag 
will be lessened, and they will tend to assume a straight line. Under these 
conditions any one point in the side spans of the cables will move vertically 
upwards, and carry the roadway up with them. In the finished bridge this 
same vertical movement, in either direction, will be affected by changes in 
the cable lengths, due to expansion and contraction. It will also occur if 
the live load coming upon the bridge is unequally distributed. Therefore, 
to maintain the relation between the ends of the side span roadways and 
the ends of the approach roadways, the cables are "tied down" at the two 
pylons, S-l and N-l. 


This plate illustrates the structural details of the Tie-Downs, one of 
which is placed on each of the cables where they pass through the pylons 
S-l and N-l, making a total of 4 altogether. Immediately following the 
squeezing operation, a collar is bolted around each cable at the point where 
it passes through the pylon. A set of 6 steel wire ropes, whose ends are 
socketed, are looped around these collars, and their ends dropped down so 
that the two parts of the ropes straddle the fixed portion of the tie-down. 
Steel rods are screwed into the lower ends of the sockets, and their lower 
ends straddle the movable portion of the tie-down. A nut screwed onto the 
lower ends of the rods fastens them to this movable portion. Just above 
the sockets a set of steel brackets are riveted to the fixed portion, against 
which the sockets bear. Flat steel shims are interposed between the sockets 
and brackets for adjustment purposes. As explained in the paragraph 
above, the placing of the steel for the roadway will cause a vertical move- 
ment in the cables at the pylons. This vertical movement of the cables at 
this point causes the sockets to bear against the brackets, which in turn will 
stop the vertical movement of the cable at the "fixed point" over the pylon. 


As the vertical movement of the cables continues between here and the 
tower, it will cause them to bend, or change their slope, at this fixed point. 
If the side span roadways do not come into perfect alignment with the ap- 
proach roadways when the entire weight of the roadway has been sus- 
pended from the cables, a set of powerful jacks is placed in between the 
movable and fixed parts of the tie-downs. As pressure is exerted by the 
jacks, the rope sockets will be pulled down away from the brackets, and 
the proper amount of shims can be added or taken out. When the pressure 
on the jacks is released, the sockets will again come to a bearing against 
the brackets, but the fixed point of the cable will be raised or lowered by 
the amount of shims that were either taken out or added. This fixed point 
is very accurately calculated, and this jacking arrangement is only pro- 
vided so that an adjustment can be made in case the actual fixed point does 
not coincide with the calculated fixed point. Bronze shoes on the ends of the 
collars, bearing against steel plates, prevent side sway at the fixed points. 



Following the squeezing operation, the cables are ready to receive the 
suspended load. The steel girders that form the supports for the roadway 
are suspended from the main cables by steel wire ropes, 2-11/16" in diame- 
ter, which are known as Suspender Ropes. These suspender ropes are af- 
fixed to the main cables by placing them over suspender rope saddles, or 
Cable Bands. These cable bands are made from cast steel, are manufactured 
in two halves, and the two halves are bolted around the cables, as shown 
in Pig. 1. Near the middle of the center span, the curve of the main cables 
is quite flat and nearly approaches the horizontal, while near the towers 
the curve becomes quite steep. The grooves in the cable bands into which 
the suspender ropes fit are cast at such an angle that the ropes will hang 
vertically when in their normal position ; the angles of the grooves varying 
as the steepness of the cable varies. Progressing toward the center of the 
span, each set of ropes is shorter than the preceding set. These cable bands 
are spaced 50 feet apart, center to center, there being 83 on each cable in 
the center span, and 22 on each side span; bringing the total number of 
bands for the two cables to 254. 


The suspender ropes are manufactured on rope making machines that 
exert a certain amount of tension on the wire as it is being ' 'twisted" into 
the form of rope. The more tension applied, the more closely will the wires 
be compacted, but there is a limit to the compactness obtainable. There- 
fore, when a load is applied to this type of rope the wires will seat them- 
selves more closely together, and the rope will stretch a certain amount. To 
overcome this stretch, and to insure that these ropes will support the road- 
way at its proper height, the manufacturers developed a prestressing 
machine to place the ropes under tension prior to erection. The tension 
applied is 50% greater than the normal working load. This tension is held 
for a certain period of time and then reduced to the normal working load, 
and while under this latter tension, proper corrections are made for tempera- 
ture changes, and the rope is then measured to its exact length and marked. 
The tension is then released, the rope cut at the mark, and a socket placed 
on each end. It is then placed on reels, taken to the bridge site and hoisted 
to the tower tops. The two ends of each rope are then dragged down the 
footbridge so that its middle is looped around the cable, as shown in Fig. 
2. The ends are led over two wheels of a platform mounted on the cable, 
and a bridle rope attached to it near the loop so that it can be eased down 
into position on the cable band. Two loops, or a total of 4 parts, are placed 
on each cable band, and the 4 parts held together underneath the cable by 
means of clamps. These clamps are prevented from slipping down the ropes 
by zinc buttons. 


To make the cable wrapping water-tight, a groove is provided at the 
edge of the cable band that is caulked with lead wool in the same manner 
that a plumber caulks the joints of soil pipe. The ropes supporting the foot- 
bridges are not used for suspender ropes on this bridge, and are simply cut 
away in this drawing in order to show more detail. The cables and sus- 
pender ropes are made accessible for inspecting and painting, by stringing 
a permanent set of hand ropes above the cable, as shown here. A short 
section of the roadway is also shown here to illustrate how it is supported 
at the lower ends of the suspender ropes, although neither handrail nor 
roadway would be in position at the stage of erection here described. 


4 V' 


After all the suspender ropes have been dropped down from the cable 
bands, as described in the preceding plate, a line projected along their lower 
ends will indicate the position that is to be occupied by the top members 
of the roadway structure, for it is from the lower ends of these ropes that 
the steel work is suspended. The steel for this structure was fabricated 
in the east and shipped out to the contractor's local plant by boat. From the 
plant it is barged out to the bridge site, and landed at the bases of the two 
towers. Derricks are mounted at the floor level of each tower for the pur- 
pose of hoisting the steel to this level, and these derricks remain stationary. 
Another set of derricks, known as traveler derricks, swing the steel mem- 
bers out into position ; starting at the towers, and progressing away from 
them as the steel is erected. There are four of these travelers, two that 
progress toward each other in the center span, and one each in the side 
spans that progress toward the anchorages. 


Beginning at the towers, the stiffening trusses are erected first, one 
member at a time. These trusses consist of sections or panels, each of 
which is 25 feet in length, and which terminate in a post. Each alternate 
post measures off a distance of 50 feet, that coincides with the 50 foot spac- 
ing of the suspender ropes, and as this post is placed in position, its upper 
end is hung from the suspender ropes. At the top of the post is a set of 
vertical angle irons into which the suspender ropes fit, as shown in Fig. 2. 
The lower ends of these angles seat on the rope sockets, and to permit of 
slight adjustments, provision is made for placing flat "U" shaped steel 
shims between the angles and sockets. When both trusses, one on either 
side, have been completed to the end of the 50 foot length, and the posts 
suspended from the ropes, the 90 foot floor beams are swung out crosswise 
of the bridge, and their ends fastened to the posts of the stiffening trusses. 
This plate shows a floor beam being swung into position by the traveler 
in the foreground. The diagonal members, forming the wind bracing is 
then fastened in place. On top of these floor beams are placed steel beams, 
known as stringers, and which run parallel with the axis of the bridge for 
its entire length. Only enough of these stringers are placed at this time 
to form a support for the derricks, and a track for the push car on which 
the steel members are brought out along the erected steel from the towers. 
The remainder of the stringers are placed later. The derrick is then moved 
out over the completed panels, and this procedure is repeated for each pair 
of panels. 


When the two derricks in the center span, traveling towards each 
other, meet at the mid-span point, the last two panels are joined together 
at the top chord, or upper horizontal members, of the stiffening trusses. 
The aligning of the lower chords of the trusses is accomplished by means of 
powerful hydraulic jacks, that force them apart until the rivet holes in 
both members line up. This is a new development in the connecting of the 
closing panels. When the derricks, working in the side spans, arrive at the 
last panel, the ends of the latter are fastened to rocker arms. The lower 
ends of these rocker arms are pinned to pedestals that are embedded in 
the concrete of the pylon, as shown in Plate XX. During erection, the steel 
is temporarily bolted together. Riveting crews follow up behind the der- 
ricks, and permanently rivet all the steel members together as soon as 
they are in proper alignment. 





— ! 








































>->s) ),J 

"' JIJ 




' !"•"-' 

TTTvrnrTTTiTrrTTTTrnrrn) >-r* 





""' ' 


" Jl 


>>>>>!»> t n>> nil f )>})!> t' tut 

mi rrmi 


frj>u)>j)>)>> >>>>) }))„ i)>t 


Vi> "•IM>>fltlltl/lt»l„.>,IJI)lA 

,,111)11 1 •'-•■" 

r rr| l |. W ...)'irn 


The final operation in the fabrication of the cables is to wrap them 
around their circumference with a wrapping of steel wire. This serves a 
two-fold purpose; one, to maintain the cables in a circular cross section, 
and the other, to protect the wires of the cables themselves from the ele- 
ments. Incidentally, all the wires in the cables are straight and parallel; 
there is no twist or turn given to either the wires, or to the strands. This 
wrapping is independent in function to the parallel wires that make up the 
sustaining strength of the cables, and may be likened to the wrapping that 
is placed around an ordinary flexible lamp cord. A lamp cord is made up of 
a number of small wires that function by carrying the electric current from 
a base plug to the lamp, and the wrapping holds them together and excludes 
any moisture from getting to the wires. 


The weight of the suspended steel structure, as it comes upon the cables, 
causes them to take more sag in the center span, and pulls most of the sag 
out of the side spans. Since they are flexible, this change in the amount of 
sag at the center, causes the cables to assume a new curve. Were the wrap- 
ping placed around the cables too soon, this change in the curve would 
cause the wrapping wire to work loose. There is also a certain amount of 
stretch to the cables, and the cables are not wrapped until after they are 
supporting a large portion of the suspended structure. As the cable bands 
are all in place at this stage of the erection, the wrapping is applied to that 
part of the cable that lies between the bands. The machine is built of two 
halves that are bolted together, around the cable. When the wrapping has 
proceeded to the face of one of the bands, it is unbolted, lifted over the 
band, and again bolted together on the other side. 


The wrapping wire is contained in the machine on a large bobbin, not 
unlike a large spool, with the cable passing through the hole in its center. 
The wire is led from the bobbin over a small grooved wheel and onto a 
larger wheel that passes through the frame of the machine. This larger 
wheel changes the direction of the wire and feeds it onto the cable at a 
tangent to the circumference of the latter, as shown in the drawing. The 
machine is rotated about the cable by means of a large gear just behind 
the bobbin, which is driven by an electric motor. Above this motor is a 
box containing the control mechanism for starting and stopping the motor. 
The machine contains two bobbins, the other being behind the large gear. 
On the face of the machine are four plungers that are held in position by 
a spring enclosed in a housing. At the lower ends of these plungers a shoe 
is fitted with a lip on one edge. This shoe bears down on the last few turns 
of wire, and the lip bears against the edge of the last wire, so that as the 
machine is revolved about the cable these lips pull it along, or feed it up 
the cable at the rate of two diameters of wrapping wire for each revolution. 
The cable just ahead of the machine is given a coat of heavy red lead paste, 
so that as the wrapping wires are applied this forms a water-tight seal. 
The wrapping wire is brought up snug against the cable bands, and the re- 
cess left in the latter is caulked with lead wool to form a water-tight seal. 
When the bobbins become empty, they are replenished from large reels of 
wire placed on the roadway, one end of which is led up through the foot- 
bridges to the machine. The bobbins are disengaged from the rest of the ma- 
chine, the motor is reversed, and the wire reeled onto the bobbins at a higher 
rate of speed than when it is being wrapped onto the cable. 





As the traffic, which is known as the live load, passes over the bridge, 
it is obvious that it will not be equally distributed along the entire length 
of the roadway at all times. This unequal distribution of the live load, when 
some of the spans are fully loaded while others are empty, imposes the 
extreme conditions of loading to which the bridge is subject. Figs. 2 and 3 
illustrate the manner in which the structure is deflected under these ex- 
treme conditions. Temperature variations will also cause the structure to 
deflect, due to the expansion and contraction of the steel. A combination 
of extreme temperatures and loading conditions, places the severest strain 
on the structure, viz: Assuming a normal temperature of 65°, a 40° rise 
will cause a downward movement at mid-span, with loading conditions as 
shown in Fig. 2 of 10 feet. Likewise, a 40° drop will cause an upward move- 
ment at mid-span, with loading conditions as shown in Fig. 3, of 10 feet; 
or, a total vertical movement of 20 feet over an 80° temperature variation, 
and extreme loading conditions reversed as to spans. Any movement in 
the spans will also affect the vertical alignment of the towers, viz : Assume 
that the temperature is 40° below normal and one of the side spans is fully 
loaded, the remainder of the bridge being unloaded; the tower adjacent 
to this loaded side span will be deflected 22 inches shoreward at the top. 
Or, assume that the temperature is 40° above normal, and the center span 
and one side span is fully loaded, the other side span being unloaded ; the 
tower adjacent to the empty span will be deflected 18 inches towards the 
center span. 

Contrary to general belief, the extreme length of the span has a "damp- 
ening" effect on side sway. This side sway, due to wind pressure, causes 
the suspender ropes to act as pendulums, that describe arcs as they swing 
out of vertical. For the same amount of sway, the chord of the arc would 
be approximately as great for a long rope near the towers as it would be 
for a short rope near mid-span. However, the short rope would be inclined 
at a greater angle from the vertical, and the side sway in it would im- 
mediately react as a direct pull on the cable. It would take a much greater 
movement of a long rope to effect the same angle with the vertical. Since 
the suspended structure is fastened at the towers, this greater retarding 
effect of the short ropes is applied where it is needed most, at mid-span. 
The effect of side sway due to wind pressure is shown, greatly exaggerated, 
in Fig. 4. Based on a 120 mile an hour wind, the mid-span point is capable 
of a side sway of 21 feet in either direction, or a total of 42 feet. 

Fig. 1 illustrates the details of the expansion joints at the tower con- 
nections. The suspended structure curves upwards, forming, a "camber" 
of 20 feet at mid-span, as indicated in Fig. 5. Expansion due to a rise in 
temperature, as well as the loading of the structure, will cause this curve 
to flatten out, and produce end movement at the expansion joints. Provi- 
sion is made for this by mounting the lower ends of the stiffening trusses 
on rocker arms, and permitting the pin to slide in a slot at the pin connec- 
tion. Side sway will cause the structure to pivot about the pin connection. 
To permit of free play for any or all of these movements, a gap is left at 
the end of the structure where it passes onto the steel work of the tower. 
This gap is bridged over by a grating like arrangement of steel fingers, 
the two rows of which slide into one another, forming a flexible joint for the 
smooth passage of vehicles across this gap. The function of the stiffening 
trusses is to distribute the weight of concentrated loads over a wide area 
so as to eliminate any "waves" they might cause in the roadway. 



On the shores of one of the lakes, in San Francisco's famous Golden 
Gate Park, is located a bare archway that once adorned the entrance to a 
private home, long since destroyed by fire. This archway is popularly called 
the "Portals of the Past." As this relic of a bygone day is symbolic of that 
which is past, well might the vista presented through the towers of the 
Golden Gate Bridge, as pictured in this plate, be called the "Portals of the 
Future." Two hundred forty-six feet above the waters of the Golden Gate, 
where the roadway intersects the towers, a person passing through these 
magnificent steel structures may have a view of the broad expanse of the 
Pacific to the west, and an unmatchable glimpse of the inland waters of San 
Francisco Bay on the east. Straight ahead lie the hills on the north shore 
of the Golden Gate; beyond, an empire that awaits only the opening of 
traffic across this giant among bridges to become an empire in fact, as well 
as in name. 


The towers consist of two legs, or posts, on top of which the saddles 
support the cables. If two bare columns were erected to function in this man- 
ner, the cables would brace them longitudinally, but they would be very un- 
stable in a transverse direction. Therefore, a structure of this kind is cross 
braced in such a manner as to cause it to act as a frame, or bent. In a great 
many of the older bridges this bracing was designed in the form of crosses, 
or "X" Bracing as it is called. To overcome the mechanical appearance of 
such bracing, the cross bracing above the roadway in the towers of this 
bridge, was designed as horizontal struts. These struts are in the form of 
trusses, as shown in Plate VIII, but the structural steel forming them is 
covered over with ornamental steel plates. The cables of a suspension bridge 
lend themselves to a natural grace of line, but a steel tower that is built 
primarily for strength does not yield so readily to architectural treatment, 
and the pleasing effect obtained here is apparent to all. 


As illustrated in the preceding plate, the roadway sub-structure con- 
sists of a series of 90 foot beams, connected across the stiffening trusses 
that are suspended on either side from the suspender ropes. On top of these 
cross beams are placed longitudinal "stringers" that run parallel with the 
axis of the bridge. There are twelve of these stringers running the full 
length of the suspended structure, two under each traffic lane. Once these 
stringers have all been placed and riveted, the placing of the concrete is 
carried out in the same manner as is done in the placing of the flooring 
of an office building. Forms for the concrete are suspended from these 
stringers, and the concrete poured in rectangular slabs, or panels. After the 
concrete is poured, and the load comes upon the roadway, it is transferred 
through these slabs to the stringers. The entire arrangement may be likened 
to a wooden floor system in a private residence, where the flooring runs 
crosswise over the tops of the joists, and the joists in turn rest on heavy 
timbers supported by posts. The bridge is of ample strength to accommo- 
date an inter-urban rapid transit system of electric trains, that may be in- 
stalled at some future date. At present, however, provision has only been 
made for motor vehicle traffic, to accommodate which, the 60 foot road- 
way is divided into six 10-foot lanes, 3 for north bound traffic and 3 for 
south bound traffic. A 10 % foot sidewalk is to flank the roadway, on either 
side, and the railings for these sidewalks will be 2y 2 feet inside of the sus- 
pender ropes. 






The location of the bridge places both terminals in U. S. Military Res- 
ervations, namely, the Presidio on the San Francisco side, and Fort Baker 
on the Marin side. A great deal of study was required to plan the approaches 
so as not to interfere unduly with the military works in both reservations. 
Particularly was this true on the San Francisco side where the bridge liter- 
ally lands right on top of a fort. After passing through the Presidio, the ap- 
proaches enter the City of San Francisco, and the problem then becomes 
one of co-ordinating the flow of bridge traffic with present and future ar- 
terials. Beyond the city, as well as beyond Fort Baker, to the north, the 
problem becomes one for the state system of highways. Therefore, the 
planning of the approaches and connecting roadways was worked out jointly 
by the U. S. Army Engineers, the California Highway Commission, the City 
and County of San Francisco, and the Bridge District. 

For years, the only through highway leading north from Sausalito 
passed over the Corte Madera grade, a narrow, winding roadway that has 
recently been replaced by a splendid new highway, that is practically a 
straight-away as far north as San Rafael. As this new highway leads out 
of Sausalito, it turns northward at Waldo Point. The highway leading north 
from the bridge approach, passes through the hills back of Sausalito and 
connects with this new highway at Waldo Point, 3% miles from the Bridge- 
head. A side road leading direct into Sausalito takes off just before reaching 
the short tunnel, as shown on the map. The final plans may call for a slight 
variation in the take off of this road as shown here. 

The approaches to the south of the bridge branch out into two roadways 
just beyond the toll plaza, one passing through the Presidio to the Marina 
district of San Francisco ; the other bearing south to connect with Funston 
Avenue. The approach to the Marina includes two viaducts, one called the 
high viaduct, built of steel trusses, and the other, or low viaduct, which is 
constructed of reinforced concrete. This route leads to the downtown dis- 
trict of San Francisco by way of Van Ness Avenue, or to the commercial 
district along the waterfront. The Funston Avenue route is to connect with 
19th Avenue and Sunset Boulevard, across Golden Gate Park. Future de- 
velopment anticipates a connection with Arguello Boulevard and a new 
diagonal boulevard leading into Divisadero Street. 

Under the present program of highway building, a considerable portion 
of which is already completed, there will be an entirely new route of north 
and south travel opened up when the Golden Gate Bridge is placed into 
service. It will provide a direct coast route from points south, through the 
Redwood Empire, and on into the State of Oregon, thus diverting a con- 
siderable portion of traffic that now passes through the central valleys of 
California. Daily local traffic will be stimulated and a new outlet provided 
for week-end and holiday trips of the San Francisco motorist. A new route 
to the Sacramento and Napa valleys will also be made available. 

One of the items that adds considerably to the cost of a bridge of this 
nature is the matter of approaches. Where a bridge lands in a densely 
populated city, it is necessary to purchase numerous pieces of property, 
some of which can only be procured by condemnation proceedings. In the 
case of the Golden Gate Bridge, both approaches are located in military 
reservations, and no purchase of property was required, although the Bridge 
District was obligated to defray the cost of moving several military struc- 


I I 


l'. \ 







This plate is an illustration of the bridge as it will appear when com- 
pleted, looking in a northwesterly direction. In the foreground is the road- 
way along the Presidio shore, used during the construction period to trans- 
port materials out to the bridge site. At the end of this roadway is located 
Fort Point, the old brick fort which the military authorities decreed should 
remain intact. The roadway of the bridge is carried over the top of this 
fort by a steel arch span, 320 feet long and 185 feet high. This steel arch is 
sprung between the two pylons, S-l and S-2. Beyond pylon S-2, to the left 
in the drawing, is the housing for the south anchorage. The location and 
design of the anchorages, was governed by the topography of the land, and 
the necessity for carrying the bridge structure clear of the fort. 


The side span terminates at the point where the cables enter pylon S-l, 
and are fastened to the tie-downs. This point, as well as the corresponding 
point at the end of the Marin side span, is 223 feet above sea level, and the 
bridge is symmetrical between these two points. The elvation of the anchor- 
age, taken at a point where the center line of the cables intersects the center 
line of the strand shoe pins, is, for the San Francisco anchorage, 40 feet 
above sea level, and 583 feet back from the end of the side span. For the 
Marin anchorage, this corresponding point is 154% feet above sea level. 
A point on the cable at this elevation falls 211 feet back from the end of the 
side span, which places it between the two pylons N-l and N-2. Being built 
on a bare hillside, and there being no structures to interfere, the anchorage 
blocks were located between the two pylons — a different arrangement than 
that for the San Francisco anchorage, shown in this plate. ( See Plate II for 
scale drawing, showing these differences) . 


The first regular suspension bridge was built in 1796 under a U. S. pat- 
ent granted to James Finlay, and between 1796 and 1810, fifty bridges were 
built under this patent. The supporting cables consisted of hand forged 
chains, the largest of which consisted of a 306 foot span across the Schuyl- 
kill River at Philadelphia. The next progressive step in suspension bridges 
beyond this use of forged chains is inseparably linked with the name of 
Roebling. John A. Roebling was born in Germany, and graduated from the 
Royal University of Berlin with an engineering degree. He emigrated to the 
United States in 1831 and took up farming in western Pennsylvania. That 
decade antedated the railroad, and canal transportation was then consid- 
ered the acme of perfection. Even in those days water could not be made 
to run up hill, and where a canal passed over a mountain, the boats were 
placed on an inclined track and hauled to the top by means of hemp ropes. 
From a German periodical Roebling learned that rope was being made of 
wire in that country, and with the idea of supplanting this hemp rope for 
canal boats, he experimented on his own account. His wire rope became suc- 
cessful, and its manufacture soon became more profitable than farming. 
These canals were carried over natural streams on bridges that were limited 
as to length, and spring freshets brought disaster as the closely spaced 
piers were set right in the stream. Roebling believed that these bridges 
could be suspended from his wire rope with a great enough span to permit 
the piers to clear the water. His first commission was for a bridge over the 
Allegheny River at Pittsburgh, and its success brought him many more 



The brief historical sketch of John A. Roebling given on page 61, is here 
continued. It is altogether fitting that the same firm of John A. Roebling 
Son's Company should have continued down on to the present day, to build 
the greatest of all bridges, the Golden Gate Bridge. The original John A. 
Roebling built a wire cable suspension bridge over the Niagara Falls rapids 
in 1854, and another over the Ohio River at Cincinnati that was opened for 
traffic in 1867. While building the Brooklyn Bridge, 1870-83, he received in- 
juries that cost him his life, and this famous structure was completed by 
his son, Col. Washington A. Roebling. The towers were built of masonry, 
and wire made from steel was used for the first time in the cables, charcoal 
iron being the material used up to that time. Another new departure was 
the use of zinc galvanizing as a protective coating for the wire. The four 
cables, each 15%" in diameter, were made up of 19 strands, and each strand 
contained 282 wires, 0.187" in diameter, or a total of 5,358 wires for each 
cable. The total length of the cables from anchorage to anchorage is 3,578 
feet, and up to about 10 years ago, remained the longest cables in existence. 
Although carrying a much greater load than it was designed for, this bridge 
is still in service. 


The Camden bridge at Philadelphia, 1926, is suspended from two cables, 
each 30" in diameter, and were made up of 61 strands, each strand contain- 
ing 306 wires, 0.195" in diameter, or a total of 18,666 wires. The cellular 
plate steel structure of the towers, and the increased length of clear span 
between the towers places it one step beyond any previously built bridge. 
Another new departure is the suspension of the roadway from the bottoms 
of the stiffening trusses, with an overhang on the outside to accommodate 
interurban trains. The sidewalks are placed on brackets, projecting out 
from the top of the truss. 

Many new features of both design and construction are incorporated 
in the George Washington Bridge, New York, 1931. While the towers revert 
back to the lattice work steel type, many new features mark their design. 
„ The roadway is suspended without stiffening trusses, and at some future 
date when traffic warrants it, another deck will be suspended below the 
present roadway, and the connecting steel between the two decks will form 
the stiffening trusses, and bring the channel clearance above water down to 
220 feet. The cables, 4 in number, are 36" in diameter, and are made up of 
61 strands. Each strand contains 434 wires, 0.196" in diameter, making a 
total of 26,474 wires for each cable. Prior to the building of the Golden 
Gate Bridge it held the record for height of towers and length of span. 


The many new features incorporated in the design and building of the 
Golden Gate Bridge have already been given. The comparative drawing op- 
posite gives the principal dimensions of all four bridges, the towers being 
drawn to a scale of 200 feet to the inch, and the spans to a scale of 1000 
feet to the inch ; the cable cross-sections are drawn to a scale of 1/16" to 
the foot. The cable length in the center is drawn to an arbitrary scale and 
angle, and indicates the manner in which the cables on the Golden Gate 
Bridge are splayed out at the anchorages. The Golden Gate Bridge is pre- 
eminently the largest single span suspension bridge in the world, but it is 
futile to predict that its size will not be exceeded. Such a bridge, to span the 
lower reaches of New York Harbor, is already under discussion. 



The first step taken in the actual building of the Golden Gate Bridge 
was the passage of an act by the Legislature of the State of California, 
providing for the creation of the bridge district. The supervisors of the 
various counties interested passed resolutions of intention to join the dis- 
trict, and 10 per cent of the voters were required to ratify these resolutions 
by petition. After the petitions were verified by the county clerks and by the 
secretary of state, this latter official then issued a certificate of incorpora- 
tion. The county supervisors then appointed their alloted number of repre- 
sentatives to serve on the Board of Directors of the District. As organized, it 
is known as the "Golden Gate Bridge and Highway District", and includes 
all of the City and County of San Francisco, the Counties of Marin, Sonoma, 
and Del Norte, and some 98% of Napa and 32% of Mendocino Counties. As 
a preliminary, the Board of Directors proceeded to organize, and levied a 
tax to defray the expense of preparing plans and estimates of cost. This 
estimate, not to exceed $35,000,000, together with the plan of the project, 
was submitted to the voters of the District. Under the requirements of the 
law for a two-thirds majority of all the voters, the bond issue was carried, 
authorizing the District to issue bonds up to the $35,000,000 limit. The 
limitations were that the amount of the bond issue would cover the entire 
cost of the bridge, including engineering fees and interest charges during 
construction, but was not to exceed 15% of the assessed valuation of the 
district. Also, that the maximum life of the bonds is to be 40 years, with a 
provision for a maximum interest rate, since fixed at a rate not to exceed 
5% per annum. 

To arrive at the cost of construction, estimates were prepared by the 
engineers responsible for the design and construction of the bridge, and 
these estimates were independently checked by a separate group of consult- 
ing engineers. In addition to the actual cost of the structure, approaches 
and plazas complete, there were included in the estimate, the items of en- 
gineering and general expenses, interest on the bonds during construction, 
and for the first six months period of operation. The cost of operating the 
bridge will include such items as administration, employees payroll, main- 
tenance, lighting, and repainting of the structure every four years. There 
will be no income tax chargeable to the operating income, and the revenue 
over and above this income will be applied to bond redemption and interest 
charges. An analysis of these estimates is beyond the scope of this book. 
Based on contracts already let, plus those yet to be let, the estimated con- 
struction cost was placed at $26,715,000 in the latter part of 1935. 

The revenue is divided up into three items, namely, from vehicles, rapid 
transit passengers, and concessions. A great deal of study has been given to 
the amount of toll to be charged, and this amount will be governed by a num- 
ber of factors yet to be determined. However, a preliminary survey and 
study by the traffic engineers produced a set of schedules that will be used 
as a tentative base in the final determination. In the original report sub- 
mitted by these engineers the fact was developed that the average toll per 
vehicle in effect at that time on the San Francisco-Marin ferry lines was 
99.3c. These ferry lines will not be discontinued upon completion of the 
bridge, and, anticipating competition, one of the schedules prepared was 
somewhat lower than this. The bridge is scheduled to be thrown open for 
traffic in the early part of 1937.