TUNNELS THROUGH HARD ROCK (Continued).--EXCAVATION BY HEADINGS.
EUROPEAN AND AMERICAN METHODS.
The more common method of tunneling through hard rock is to begin the
work by a heading, instead of by a drift. This heading may be of small
dimensions, and the remainder of the section may also be removed in
successive small parts, or it may be the full width of the section, and
the enlargement of the section be made in one other cut.
[Illustration: FIG. 61.--Diagram Showing Sequence of Excavation in
Heading Method of Tunneling Rock.]
=General Discussion.=--When the tunnel is excavated by means of several
cuts, which is the method usually employed in Europe, the sequence of
work is as indicated by Fig. 61. Work is begun by driving the center top
heading No. 1, whose floor is at the level of the bottom of the roof
arch, and which is usually excavated by the circular cut method. This
heading is widened by removing parts Nos. 2 and 3 until the top part of
the section is removed, then the roof arch is built with its feet
resting on the unexcavated rock below. The lower portion of the section
or bench is removed by first sinking the trench No. 4, after which part
No. 5 is taken out, and then parts Nos. 6 and 7, and the side walls
built. Part No. 8 for the culvert is finally opened. The heading is, as
a rule, driven far in advance, but the excavation of each of the other
parts follows the preceding one at a distance behind of about 300 ft.
The strutting, when any is required, is usually the typical radial
strutting of the Belgian method of tunneling. The masonry lining is
constructed practically the same as in tunnels excavated by a drift. The
hauling is done on a single track laid in the heading No. 1, which
separates into double tracks where the full top section has been
excavated by the removal of parts No. 2. These two tracks are again
combined and form a single track along the top of part No. 5, which has
been left wider than part No. 4 for this particular purpose. When part
No. 3 is excavated a standard-gauge track is laid on its floor; and as
the full section of the tunnel is completed by taking out parts Nos. 4
and 5, this single track is replaced by two standard-gauge tracks, into
which it switches. Spoil is transferred from the narrow-gauge tracks on
the upper level, to the standard-gauge tracks on the tunnel floor, by
means of chutes, and building material is transferred in the opposite
direction by means of hoisting apparatus.
When the excavation is made by a single wide heading, and a single other
cut for removing the bench, which is the method preferred by American
engineers, it is called the Heading and Bench method. The work begins by
removing a top heading the full width of the section; this heading is
usually made 7 ft. or 8 ft. high, and is excavated by the center cut
method. The method of strutting usually employed is to erect successive
three- or five-segment timber arches, whose feet rest on the top of the
bench; when the bench is removed, posts are inserted under the feet of
each arch. These arches are covered with a lagging of plank. In America
it has often been the practice to let this strutting serve as a
temporary lining, and to replace it only after some time, often after
years, with a permanent lining of masonry. In a succeeding chapter, some
of the methods adopted in relining timber-lined arches with masonry are
described. The hauling is done by either narrow or broad gauge tracks
laid on the floor of the completed section below. A device called a
bench carriage is often employed to enable the cars running on the
heading tracks to dump their loads into the cars below, without
interfering with the work on the bench front. This device consists of a
wide platform carried on trucks, running on rails at the sides of the
tunnel floor, so that it is level with the floor of the heading. The
front of this platform carries a hinged leaf which may be raised and
lowered, and which forms a sort of gang-plank reaching to the floor of
the heading. By running the heading cars out on to this traveling
platform, they can be dumped into the cars below entirely clear of the
work in progress on the bench front.
For the purpose of illustrating the two methods of driving tunnels by a
heading, which have been briefly described, the St. Gothard and the Fort
George tunnels have been selected. The St. Gothard tunnel is selected,
as being one of the longest tunnels in the world, and because it was
excavated by a number of small parts; and the Fort George tunnel, as
being a double-track tunnel, driven by a heading, and bench, and having
a concrete lining.
ST. GOTHARD TUNNEL.
The St. Gothard tunnel penetrates the Alps between Italy and France, and
is 9¹⁄₄ miles long. It was constructed in 1872-82.
=Material Penetrated.=--The St. Gothard tunnel was excavated through
rock, consisting chiefly of gneiss, mica-schist, serpentine, and
hornblende, the strata having an inclination of from 45° to 90°. At many
points the rock was fissured, and disintegrated easily, and water was
encountered in large quantities, causing much trouble.
=Excavation.=--The sequence of excavation is shown by Fig. 14, p. 36.
First the top center heading, No. 1, whose dimensions varied from 8.25 ×
8.6 ft. to 8.5 × 9 ft., according to the quality of the rock, was driven
never less than 1000 ft. and sometimes over 3000 ft. in advance of parts
No. 2. The excavation of parts No. 2 opened up the full top section, and
parts Nos. 3, 4, 5, 6, and 7, were removed in the order numbered.
=Strutting.=--Where regular strutting was required, the construction
shown in Fig. 62 was adopted.
=Masonry.=--The St. Gothard tunnel is lined throughout with masonry.
After the upper portion of the section was fully excavated, the roof
arch was built with its feet resting upon short planks on the top of the
bench. Plank centers were used in constructing the arch. For the arch
brick masonry was employed, but the side walls were built of rubble
masonry. Shelter niches, about 3 ft. deep, were built into the side
walls at intervals, and about every 3,000 ft. storage niches about 10
ft. deep, and closed with a door, were constructed. The culvert was of
brick masonry.
=Mechanical Installation.=--Water-power was used exclusively in driving
the St. Gothard tunnel. At the north end, the Reuss, and at the south
end, the Tessin and the Tremola, rivers or torrents were dammed, and
their waters conducted to turbine plants at the opposite ends of the
tunnel. The power thus furnished by the Reuss was about 1,500 H.P., and
the power furnished by the combined supply of the Tessin and Tremola was
1,220 H.P. The turbine plant at both ends at first consisted of four
horizontal impulse turbines, but later, two more turbines were added at
the south end. Each of the two sets of four turbines first installed
drove five groups of three compressors each, and the two supplementary
turbines drove two groups of four compressors each. The compressors were
of the Colladon type with water injection, and four groups of three
compressors each were capable of furnishing 1,000 cu. yds. of air
compressed to between seven and eight atmospheres every hour, or about
100 H.P. per hour, delivered to the drills at the front. This air when
exhausted provided about 8,000 cu. yds. of fresh air per hour for
ventilation.
The compressors at each entrance discharged into a group of four
cylindrical receivers of wrought-iron each 5.3 ft. in diameter by 29.5
ft. long, and having a capacity of 593 cu. ft. The cylinders were placed
horizontally, the first one receiving the air at one end and discharging
it at the other end into the next cylinder, and so on. By this
arrangement the air was drained of its moisture, and the discharge from
the end receiver into the tunnel delivery pipes was not affected by the
pulsations of the compressors. The delivery pipe decreased from 8 in. in
diameter at the receiver to 4 ins. in diameter, and finally to 2¹⁄₂ ins.
in diameter, at the front.
The drills employed were of various patterns. The first one employed was
the Dubois & François “perforator,” in which the drill-bit was fed
forward by hand. This was replaced by Ferroux drills having an automatic
feed. Jules McKean’s “perforator” was employed at the north end of the
tunnel. All of these drills were of the percussion type, and were
mounted on carriages running on tracks. Their comparative efficiency was
officially tested in drilling granitic gneiss with an operating air
pressure of 5.5 atmospheres with the following results:
NAME OF DRILL. PENETRATION
INS. PER MIN.
Ferroux 1.6
McKean 1.4
Dubois & François 1.04
Soummelier 0.85
The heading was excavated by the circular cut method, the holes being
driven as follows: Near the center of the heading three holes were first
drilled, converging so as to inclose a pyramid with a triangular base.
Around these center holes from 9 to 13 others were driven parallel to
the tunnel axis. The center holes were blasted first, and then the
surrounding holes. From 3 to 5 hours were required to drill the two sets
of holes, and from three to four hours were required to remove the
blasted rock. The number of holes drilled in removing each of the
various parts was as follows:
Part No. 1 6 to 9
Part No. 2 6 to 10
Part No. 3 2
Part No. 4 6 to 9
Part No. 5 3
Part No. 6 6 to 9
Part No. 7 1
--------
Total for full section 36 to 40
=Hauling.=--Two different systems were employed for hauling the spoil
and construction material in the St. Gothard tunnel. To remove the spoil
from parts Nos. 1 and 2 a narrow-gauge track was laid on the floor of
the heading, and the cars were hauled by horses, the grade being
descending from the fronts. These narrow-gauge cars were dumped into
larger broad-gauge cars running on the track laid on the floor of the
completed section and hauled by compressed air locomotives (Fig. 63). To
raise the incoming structural material from the broad-gauge cars to the
narrow-gauge cars running on the level above, hoisting devices were
employed.
[Illustration: FIG. 62.--Method of Strutting Roof, St. Gothard Tunnel.]
[Illustration: FIG. 63.--Sketch Showing Arrangement of Car Tracks, St.
Gothard Tunnel.]
FORT GEORGE TUNNEL.[10]
From a point north of 157th Street and Broadway almost to Dyckman
Street, that is, a distance of nearly two miles, the New York Subway
passes under an elevation known as Fort Washington Heights, which almost
bounds Manhattan Island at its upper end near the Harlem Ship Canal.
Under this elevation the rapid transit railroad was constructed in
tunnel. The tunnel was driven from two intermediate shafts over 110 ft.
deep, located one at 169th Street and the other at 181st Street and
Broadway. Both shafts were sunk at one side of the center line of the
tunnel. After these shafts had been utilized for working purposes during
the construction of the tunnel, they were equipped with electric
elevators to carry passengers from the streets to the deep station.
[10] Condensed from a paper by Stephen W. Hopkins in _Harvard
Engineering Journal_, April, ’08.
=Material.=--The material encountered in the excavation of the Fort
George tunnel was the usual mica schist met everywhere on Manhattan
Island. It was full of seams with strata running in every direction to
such an extent that at many points the roof of the tunnel had to be
supported by timbers; at other parts along the line the rock was so
disintegrated that it was considered a very loose and treacherous soil.
Two serious accidents, each accompanied by loss of life, occurred during
the construction of this tunnel. Both of them were caused by the sudden
fall of a large ledge of rock which, after the tunnel had been excavated
to the full section, remained hanging on the roof, deprived of any
support and held in place by the little cohesion of the material packing
the seams.
=Excavation.=--The tunnel was excavated by the heading method in only
two cuts, viz., the heading and bench as indicated in the Fig. 65. The
heading, almost as wide as the upper portion of the tunnel section, was
excavated in the manner explained on page 91. After the heading was
removed, the enlargement of the entire upper section of the tunnel was
accomplished by driving three inclined holes at each side of the
heading. They were driven at different depths and inclinations, as shown
in the figure and were called trimming holes. At the same time the bench
was removed by means of five holes--three vertical and two inclined. The
line of subgrade was reached by means of five grading holes driven
almost horizontal with a slight inclination downward. The air drills for
the heading were mounted on columns, all the others on tripods. The
blasting was done in the following order: the grading holes were blasted
in the first round, the bench and trimming in the second, the center cut
of the heading in the third, the sides in the fourth and the dry holes
in the last. Thus each advance of 7 ft. of the whole tunnel section was
made by means of forty holes fired in five rounds which consumed 277
lbs. of dynamite with an average additional quantity of 76 lbs., making
a total of 353 lbs. With the exception of the center cut, where 60%
dynamite was used, all the other holes were discharged with 40%
dynamite.
[Illustration: ~Cross Section.~
~Longitudinal Section.~
FIG. 64.--Arrangement of Drill Holes in the Fort George Tunnel.
FIG. 65.--Longitudinal Section of the Heading and Bench Excavation at
the Fort George Tunnel.]
=Strutting.=--When the rock was of such a character as to be dangerous
and required permanent timber support, until the masonry lining was in
place, the method employed was as follows: a top heading was first
excavated about 10 ft. deep and from 10 ft. to 12 ft. wide for some
distance, 100 ft. to 500 ft., the dangerous rock being supported by 10 ×
10 in. yellow pine plumb or raking posts and sometimes by timber bents
(“caps and legs”). The next process was to widen the heading to the full
width of 30 ft. for a length of about 20 ft., placing timber supports
under the dangerous rock as the widening-out progressed. The excavation
was deepened a little at the sides to 9.5 ft. below the roof grade
(ordered line of excavation) or about 11 ft. below the roof grade, which
was necessary when segmental timbering was to be used, to allow for
placing a 12 × 12 in. “wall plate” (timber sill) along each side. These
wall plates, generally 20 ft. long, were set to the correct elevation
and were leveled by blocking and wedging. As soon as the wall plates
were set, the work of erecting the segmental timber sets, one set at a
time, was begun by starting from the wall plates and supporting the
timber on scaffolding until keyed in, then it was blocked up to the rock
at each joint and at other necessary points. When two or more sets were
erected, lagging, made of boards 2 ins. thick by 6 to 10 ins. wide, was
placed over the segmental timber “sets” and the space above the timber
dry packed with small stone placed by hand. Sometimes there was enough
room between the timber and the rock to do all the dry packing after the
full number of sets, generally six, had been placed on the two wall
plates. The temporary timber posts and braces were taken out as the
segmental timber sets were erected.
The seven timbers that made up a timber set were of yellow pine each
10 × 10 ins., 5 ft. 2 ins. long at the intrados and 5 ft. 6 ins. at the
extrados. The sets were spaced from 3 ft. to 5 ft. apart, but generally
3.5 ft. and braced to each other at each joint of the segmental timbers
by 6 × 8 in. spreaders which were wedged against the joint splices.
When the timbers were all erected on a set of wall plates (20 ft.) and
the lagging and dry packing were completed the work of taking out the
bench, which had been partly drilled as the timber sets were erected,
was resumed. The face of the bench, which had been left about 4 ft. from
the end of the previous set of wall plates, was brought forward slowly
by placing 10 × 10 in. plumb posts which extended below subgrade under
the wall plates. These posts were generally spaced the same as the
timber sets above and directly under them.
When the face of the bench had been brought to within 3 or 4 ft. of the
forward end of the wall plate, the process of widening out and timbering
another 20 ft. length of heading was begun. In some places the rock,
though needing permanent support, was such that the work of taking out
the bench and widening the heading was carried on simultaneously without
increasing the danger; but the greater portion of the work, when
strutting was required, was done as has been described.
[Illustration: FIG. 66.--Diagram Showing the Arrangement of Drill Holes
in the Heading and Bench of the Gallitsin Tunnel.]
=Hauling.=--The excavated material was loaded at the foot of the bench
in dump cars which were run by mule power to the portal or the shaft
according to location, on 36 in. gauge-service tracks. Inclines at 159th
Street were graded from the portal at 158th Street to the street
surface. The cars were formed at this portal into a train and were taken
up the incline to the dump at 162nd Street and the North River by
construction locomotives. At the 168th Street and 181st Street shafts,
the cars were hoisted to the surface in cages (elevators). In the former
case, they were taken to the dump at 165th Street and the North River by
mules and gravity; in the latter case, to various dumps by teams. At
both shafts, stone crushers were located, therefore a great part of the
material did not have to be hauled to the dumps or even taken to the
surface as a great deal of stone was used in dry packing over the
concrete arch. The material from the portal at Fort George was hauled by
mules directly to the dump near by.
=Lining.=--The entire tunnel was lined with concrete, consisting of a
floor 4 ins. thick and vertical side walls 18 ins. thick and 25 ft.
apart, which carried a semicircular arch 18 ins. thick except in the
timbered portions where the thickness was increased to 21 ins. and to 24
and 27 ins. in some places. The springing line of the arch is 6 ft. 2
ins. above the concrete floor (5 ft. 6 ins. above the base of rail),
hence the maximum clearance above the base of rail is 18 ft. The side
walls and arch were built solid of rock to a height of 8 ft. above
springing line and the space above that point between the concrete and
the rock was packed by hand with small stones. The concrete of the arch
was laid on timber centers erected for that purpose.
The heading and bench method of excavating rock tunnels is not always
followed in the manner just described but is employed with slight
modifications. There is a large variety of modifications but only the
two most commonly used in practical works are given here. The heading
and bench method illustrated in Fig. 66 was used, among others, on the
Gallitsin tunnel along the Pennsylvania R.R. at the summit of the
Alleghenies near Altoona, Pa., and more recently in the tunnels
constructed by the same company under Bergen Hill, N. J., for the
entrance to New York City. The shape of the cross-section of these
tunnels was semicircular arch on vertical side walls. The excavation was
made in three consecutive cuts, viz., the heading marked 1 in the
figure, the top bench 2, and the lower bench 3. A heading 7 ft. high and
10 ft. wide was attached near the crown of the arch and the rock was
removed by means of a center cut and parallel side holes, the number of
holes depending upon the consistency of the rock. The part No. 2 was
excavated by drilling holes at each side to different depths and at
different inclinations in order to reach the line of the profile as well
as the springing line of the proposed tunnel. The central part of the
top bench was excavated by means of holes driven vertically from the
floor of the heading. The bottom bench No. 3, included between the
springing line of the arch and subgrade, was removed by means of five
vertical holes driven from the floor of the top bench. The three
different working parts were kept nearly 10 ft. apart. Blasting was
effected in reversed order to the figures marked in the diagram, viz.,
the bottom bench first and the heading last.
[Illustration: FIG. 67.--Diagram Showing a Modification of the Heading
and Bench Method.]
Still another modification of the heading and bench method, commonly
followed by American engineers, is the one shown in Fig. 67. This
consists in dividing the tunnel section in three parts by horizontal
lines. The resultant parts are first the heading excavated close to the
roof, and as wide as the whole section of the tunnel; second, the top
bench in the middle, and lastly the bottom bench excavated to the depth
of the proposed tunnel floor. The excavation proceeds in the numerical
order, beginning at the heading which was excavated, as usual, by means
of a center cut and side holes to the full width of the proposed tunnel.
First the top bench, then the bottom bench, are removed by means of
vertical holes driven from the floor of the heading and the floor of the
top bench, respectively.
COMPARISON OF METHODS.
The differences between the drift and heading methods of excavating
tunnels through rock, consist chiefly in the excavations, strutting, and
hauling. When the drift method is employed an advanced gallery is opened
along the floor of the tunnel before the upper part of the section is
removed, and when the heading method is employed the upper part of the
section is completely excavated before any part of the section below is
excavated. When the drift method of driving is employed polygonal
strutting is usually used, and longitudinal strutting is employed with
the heading method of driving. In the drift method the hauling is done
by one system of tracks at the same level, while in the heading method
two systems of tracks are employed at different levels.
It is, perhaps, impossible to state without qualification which method
is the better. European engineers who have been connected with both the
Mont Cenis and St. Gothard tunnels, driven by the drift and heading
methods respectively, had the opportunity to practically observe the
advantages and disadvantages of these two methods. Their conclusion was
that the drift method was more convenient for tunnels driven through
hard and compact rock, and that the heading method was better for
tunnels of fissured and disintegrated rocks. To prove this opinion,
experiments were made in one of the tunnels approaching the great St.
Gothard tunnel. On a short tunnel the excavation was made by the drift
method from one portal, while at the other, the heading method was
followed. Although the general rule was fully confirmed still the
conditions at the portals were not identical. More conclusive
experiments were made by Mr. Ira A. Shaler, the contractor for Section
IV., of New York Rapid Transit Railway. He had the opportunity of
driving two parallel tunnels under Murray Hill only 17 ft. apart. The
eastern tunnel was driven by the drift method, the western one by the
heading method. After the work had proceeded for a few months, Mr.
Shaler stated that in his case the drift method was more convenient. He
could spare drilling several holes at each advance, thus obtaining
economy in time, labor and material without considering the advantage of
a simpler transportation of the débris. He promised to publish his
results for the benefit of the profession, but, unfortunately, lost his
life in an accident in the tunnel before the completion of the work.
An advantage that the drift method affords in long tunnels is, that the
water, which is usually found in large quantities under high mountains,
is easily collected in the drift and conveyed to the culvert, while in
the heading method the water from the advance gallery, before being
collected into the culvert built on the floor of the tunnel, must pass
through all the workings. This may be a serious inconvenience when water
is found in large quantities, as, for instance, was the case in the St.
Gothard tunnel, where the stream amounted to 57 gallons per second.