TUNNEL (Fr. tonnel, later tonneau, a diminutive from Low Lat. lonna, tunna, a tun, cask), a more or less horizontal underground passage made without removing the top soil. In former times any long tube-like passage, however constructed, was called a tunnel. At the present day the word is sometimes popularly applied to an underground passage constructed by trenching down from the surface to build the arching and then refilling with the top soil; but a passage so constructed, although indistinguishable from a tunnel when completed, is more correctly termed a " covered way," and the operations " cutting " and " covering," instead of tunnelling. Making a small tunnel, afterwards to be converted into a larger one, is called " driving a heading," and in mining operations small tunnels are termed " galleries," " driftways " and " adits." If the underground passage is vertical it is a shaft; if the shaft is begun at the surface the operations are known as "sinking"; and it is called a " rising " if worked upwards from a previously constructed heading or gallery.
Tunnelling has been effected by natural forces to a far greater extent than by man. In limestone districts innumerable swallow-holes, or shafts, have been sunk by the rain water following joints and dissolving the rock, and from the bottom of these shafts tunnels have been excavated to the sides of hills in a manner strictly analogous to the ordinary method of executing a tunnel by sinking shafts at intervals and driving headings therefrom. Many rivers find thus a course underground. In Asia Minor one of the rivers on the route of the Mersina railway extension pierces a hill by means of a natural tunnel, whilst a little south at Seleucia another river flows through a tunnel, 20 ft. wide and 23 ft. high, cut 1600 years ago through rock so hard that the chisel marks are still discernible. The Mammoth Cave of Kentucky and the Peak caves of Derbyshire are examples of natural tunnelling. Mineral springs bring up vast quantities of matter in solution. It has been estimated that the Old Well Spring at Bath has discharged since the beginning of the 1pth century solids equivalent to the excavation of a 6 ft. by 3 ft. heading 9 m. long; and yet the water is perfectly clear and the daily flow is only the isoth part of that pumped out of the great railway tunnel under the Severn. Tunnelling is also carried on to an enormous extent by the action of the sea. Where the Atlantic rollers break on the west coast of Ireland, or on the seaboard of the western Highlands of Scotland, numberless caves and tunnels have been formed in the cliffs, beside which artificial tunnelling operations appear insignificant. The most gigantic subaqueous demolition hitherto carried out by man was the blowing up in 1885 of Flood Rock, a mass about 9 acres in extent, near Long Island Sound, New York. To effect this gigantic work by a single instantaneous blast a shaft was sunk 64 ft. below sea-level, from the bottom of which 4 m. of tunnels or galleries were driven so as to completely honeycomb the rock. The roof rock ranged from 10 ft. to 24 ft. in thickness, and was supported by 467 pillars 15 ft. square; 13,286 holes, averaging 9 ft. in length and 3 ins. in diameter, were drilled in the pillars and roof. About 80,000 cub. yds. of rock were excavated in the galleries and 275,000 remained to be blasted away. The holes were charged with no tons of " rackarock," a more powerful explosive than gunpowder, which was fired by electricity, when the sea was lifted 100 ft. over the whole area of the rock. Where natural forces effect analogous results, the holes are bored and the headings driven by the chemical and mechanical action of the rain and sea, and the explosive force is obtained by the expansive action of air locked up in the fissures of the rock and compressed to many tons per square foot by impact from the waves. Artificial breakwaters have often been thus tunnelled into by the sea, the compressed air blowing out the blocks and the waves carrying away the debris.
With so many examples of natural caves and tunnels in existence it is not to be wondered at that tunnelling was one of the earliest works undertaken by man, first for dwellings and tombs, then for quarrying and mining, and finally for water-supply, drainage, and other requirements of civilization. A Theban king on ascending the throne began at once to drive the tunnel which was to form his final resting-place, and persevered with the work until death. The tomb of Mineptah at Thebes was driven at a slope for a distance of 350 ft. into the hill, when a shaft was sunk and the tunnel projected a farther length of about 300 ft., and enlarged into a chamber for the sarcophagus. Tunnelling on a large scale was also carried on at the rock temples of Nubia and of India, and the architectural features of the entrances to some of these temples might be studied with advantage by the designers of modern tunnel fronts. Flinders Petrie has traced the method of underground quarrying followed by the Egyptians opposite the Pyramids. Parallel galleries about 20 ft. square were driven into the rock and cross galleries cut, so that a hall 300 to 400 ft. wide was formed, with a roof supported by rows of pillars 20 ft. square and 20 ft. apart. Blocks of stone were removed by the workmen cutting grooves all round them, and, where the stone was not required for use, but merely had to be removed to form a gallery, the grooves were wide enough for a man to stand up in. Where granite, diorite and other hard stone had to be cut the work was done by tube drills and by saws supplied with corundum, or other hard gritty material, and water the drills leaving a core of rock exactly like that of the modern diamond drill. As instances of ancient tunnels through soft ground and requiring masonry arching, reference may be made to the vaulted drain under the south-east palace of Nimrod and to the brick arched tunnel, 12 ft. high and 15 ft. wide, under the Euphrates. In Algeria, Switzerland, and wherever the Romans went, remains of tunnels for roads, drains and water-supply are found. Pliny refers to the tunnel constructed for the drainage of Lake Fucino as the greatest public work of the time. It was by far the longest tunnel in the world, being more than 3^ m. in length, and was driven under Monte Salviano, which necessitated shafts no less than 400 ft. in depth. Forty shafts and a number of " cuniculi," or inclined galleries, were sunk, and the excavated material was drawn up in copper pails, of about ten gallons capacity, by windlasses. The tunnel was designed to be 10 ft. high by 6 ft. wide, but its actual crosssection varied. It is stated that 30,000 labourers were occupied eleven years in its construction. With modern appliances such a tunnel could be driven from the two ends without intermediate shafts in eleven months.
No practical advance was made on the tunnelling methods of the Romans until gunpowder came into use. Old engravings of mining operations early in the 17th century show that excavation was still accomplished by pickaxes or hammer and chisel, and that wood fires were lighted at the ends of the headings to split and soften the rock in advance (see fig. i).
(From Agricola's De re metallica, Basel, 1621.)
FIG. I. Method of mining, 1621.
Crude methods of ventilation by shaking cloths in the headings and by placing inclined boards at the top of the shafts are also on record. In 1766 a tunnel 9 ft. wide, 12 ft. high and 2880 yds. long was begun on the Grand Trunk Canal, England, and completed eleven years later; and this was followed by many others. On the introduction of railways tunnelling became one of the ordinary incidents of a contractor's work; probably upwards of 4000 railway tunnels have been executed.
Tunnelling under Rivers and Harbours. In 1825 Marc Isambard Brunei began, and in 1843 completed, the Thames tunnel between Rotherhithe and Wapping now used by the East London railway. He employed a peculiar " shield," made of timber, in several independent sections. Part of the ground penetrated was almost liquid mud, and the cost of the funnel was about 1300 per lineal yard. In 1818 he took out a patent for a tunnelling process, which included a shield, and which mentioned cast iron as a surrounding wall. His shield foreshadowed the modern shield, which is substituted for the ordinary timber work of the tunnel, holds up the earth of excavation, affords space within its shelter for building the permanent walls, overlaps these walls in telescope fashion, and is moved forward by pushing against their front ends. The advantages of cast-iron walls are that they have great strength in small space as soon as the segments are bolted together, and they can be caulked water-tight.
In 1830 Lord Cochrane (afterwards loth earl of Dundonald) patented the use of compressed air for shaft-sinking and tunnelling in water-bearing strata. Water under any pressure can be kept out of a subaqueous chamber or tunnel by sufficient air of a greater pressure, and men can breathe and work therein for a time up to a pressure exceeding four atmospheres. The shield and cast-iron lining invented by Brunei, and the compressed air of Cochrane, have with the aid of later inventors largely removed the difficulties of subaqueous tunnelling. Cochrane's process was used for the foundation of bridge piers, etc., comparatively early, but neither of these devices was employed for tunnelling until half a century after their invention. Two important subaqueous tunnels in the construction of which neither of these valuable aids was adopted are the Severn and the Mersey tunnels.
The Severn tunnel (fig. 16), 4$ m. in length for a double line of railway, begun in 1873 and finished in 1886, Hawkshaw, Son, Hayter & Richardson being the engineers and T. A. Walker the contractor, is made almost wholly in the Trias and Coal Measure formations, but for a short distance at its eastern end passes through gravel. At the lowest part the depth is 60 ft. at low water and looft. at high water, and the thickness of sandstone over the brickwork is 45 ft. Under a depression in the bed of the river on the English side there is a cover of only 30 ft. of marl. Much water was met with throughout. In 1879 the works were flooded for months by a land spring on the Welsh side of the river, and on another occasion from a hole in the river bed at the Salmon Pool. This hole was subsequently filled with clay and the works completed beneath. Two preliminary headings were driven across the river to test the ground.
Break-ups " were made at intervals of two to five chains and the arching was carried on at each of these points. All parts of the excavation were timbered, and the greatest amount excavated in any one week was 6000 cub. yds. The total amount of water raised at all the pumping stations is about 27,000,000 gallons in twenty-four hours.
The length of the Mersey tunnel (fig. 15) between Liverpool and Birkenhead between the pumping shafts on each side of the river is one mile. From each a drainage heading was driven through the sandstone with a rising gradient towards the centre of the river. This heading was partly bored out by a Beaumont machine to a diameter of 7 ft. 4 in. and at a rate attaining occasionally 65 lineal yds. per week. All of the tunnel excavation, amounting to 320,000 cub. yds., was got out by hand labour, since heavy blasting would have shaken the rock. The minimum cover between the top of the arch and the bed of the river is 30 ft. Pumping machinery is provided for 27,000,000 gallons per day, which is more than double the usual quantity of water. Messrs Brunlees & Fox were the engineers, and Messrs Waddell the contractors for the works, which were opened in 1886, about six years after the beginning of operations.
In 1869 P. W. Barlow and J. H. Greathead built the Tower foot-way under the Thames, using for the first time a cast-iron lining and a shield which embodied the main features of Brunei's design. Barlow had patented a shield in 1864, and A. E. Beach one in 1868. The latter was used in a short masonry tunnel under Broadway, New York City, at that time. In 1874 Greathead designed and built a shield, to be used in connexion with compressed air, for a proposed Woolwich tunnel under the Thames, but it was never used. Compressed air was first used in tunnel work by Hersent, at Antwerp, in 1879, in a small drift with a cast-iron lining.
In the same year compressed air was used for the first time in any important tunnel by D. C. Haskin in the famous first Hudson River tunnel, New York City. This was to be of two tubes, each having internal dimensions of about 16 ft. wide by 1 8 ft. high. The excavation as fast as made was lined with thin steel plates, and inside of these with brick. In June 1880 the northerly tube had reached 360 ft. from the Hoboken shaft, but a portion near the latter, not of full size, was being enlarged. Just after a change of shifts the compressed air blew a hole through the soft silt in the roof at this spot, and the water entering drowned the twenty men who were working therein. From time to time money was raised and the work advanced. Between 1888 and 1891 the northerly tunnel was extended 2000 ft. to about three-fourths of the way across, with British capital and largely under the direction of British engineers Sir Benjamin Baker and E. W. Moir. Compressed air and a shield were used, and the tunnel walls were made of bolted segments of cast iron. The money being exhausted, the tunnel was allowed to fill with water, and it so remained for ten years. Both tubes were completed in 1908.
The use of compressed air in the Hudson tunnel, and of annular shields and cast-iron lined tunnel in constructing the City & South London railway (1886 to 1890) by Greathead, became widely known and greatly influenced subaqueous and soft-ground tunnelling thereafter. The pair of tunnels for this railway from near the Monument to Stockwell, from 10 ft. 2 in. to 10 ft. 6 in. interior diameter, were constructed mostly in clay and without the use of compressed air, except for a comparatively short distance through water-bearing gravel. In this gravel a timber heading was made, through which the shield was pushed. The reported total cost was 840,000. Among the tunnels constructed after the City & South London work was well advanced, lined with cast-iron segments, and constructed by means of annular shields and the use of compressed air, were the St Clair (Joseph Hobson, engineer) from Sarnia to Port Huron, 1889-1890, through clay, and for a short distance through water-bearing gravel, 6000 ft., 1 8 ft. internal diameter; and the notable Blackwall tunnel under the Thames (Sir Alexander Binnie, engineer, and S. Pearson & Sons, contractors), through clay and 400 ft. of watersaturated gravel, 1892-1897, about 3116 ft. long, 24 ft. 3 in. in internal diameter. The shield, 19 ft. 6 in. long, contained a bulkhead with movable shutters, as foreshadowed in Baker's proposed shield (fig. 2). Numerous tunnels of small diameter have been similarly constructed under the Thames and Clyde for electric and cable ways, several for sewers in Melbourne, and two under the Seine at Paris for sewer siphons.
The Rotherhithe tunnel, under the Thames, for a roadway, with a length of 4863 ft. between portals, of which about 1400 ft. are directly under the river, has the largest crosssection of any subaqueous tube of this FIG. 2. B. Baker's pneumatic shield, type in the world (see fig. 3). It was begun in 1904 and finished in 1908, Maurice Fitzmaurice being the engineer of design and construction, and Price & Reeves the contractors. It penetrates sandy and shelly clay overlying a seam of limestone, beneath which are pebbles and loamy sand. A preliminary tunnel for exploration, 12 ft. in diameter, was driven across the river, the top being within 2 ft. of the following main tunnel. The top of the main tunnel excavation in the middle of the river was only 7 ft. from the bed of the Thames, and a temporary blanket of filled earth, usually allowed in similar cases, was prohibited owing to the close proximity of the docks. The maximum progress in one day was 12-5 ft., and the average in six days 10-4 ft. The air compressors were together capable of supplying 1,000,000 cub. ft. of air per hour.
Some tunnels of marked importance of this type to be operated solely with electric cars have been built under the East and Hudson rivers at New York. Two tubes of 15 ft. interior diameter and 4150 ft. long penetrate gneiss and gravel directly under the East River between the Battery and Brooklyn. They were begun in 1902, with Wm. B. Parsons and George S. Rice as engineers, and were finished in December 1907, under the direction of D. L. Hough of the 1zz -02 The Thames Tunnel (Brunei), 1825-1842 Glasgow Cable Subway, Clyde.
Hudson River, Morton St. a tubes.
Baker St. & Waterloo Greenwich Footway Railway, Thames. 2 tubes. Tunnel, i tube.
Harlem River. 2 tubes.
City & South London Railway, Thames. 3 tubes.
St Clair River, i tube.
Hudson River (Haskin), 1879.
fvfwsWw* Waterloo & City Rail way, Thames.
Plackwall Tunnel, Thames, i tube.
East Boston Tunnel under Harbour, i tube.
Rotherhithe, Thames, t tube Hudson and E.iht Rivers. Pennsylvania Railroad. 2 and 4 tubes.
i^B =- Battery to Brooklyn, East River. 2 tubes.
<VV"AfiS,<" rV* .XV .&'" ^.V. v ^'^ v 'e .
Detroit River Tunnel. 2 lubes.
'. s Scale of Feet 10 kiver Seine, Pari*. i tube Fie. 3. Cross Sections ol Tunnels under Kivers and Harbours.
New York Tunnel Company. They carry subway trains. In one of the blow-outs of compressed air a workman was blown through the gravel roof into the river above. He lived until the next day. Two other tubes of the same size built also through gneiss and gravel between 1905 and 1907 by the Degnon Contracting Company, with R. A. Shailer as the contractors' engineer, go from 42nd Street to Long Island City.
Four much larger tubes (see fig. 3) built in 1904 to 1909, for the Pennsylvania railroad, with Alfred Noble as chief engineer, S. Pearson & Son as contractors, and E. W. Moir as general manager, cross from 32nd and 33rd Streets to Long Island. The maximum average progress per day (one heading) for the best month's work was: rock, 4-1 ft.; rock and earth, 3-8 ft.; earth, with full sand face, 12-8 ft. The best methods of preventing blow-outs were found to consist of employing clay blankets (sometimes 25 ft. thick) on the river bed, which could be carried up to 20 ft. depth of water, and of filling the pores of the sand and gravel with blue lias lime or cement grout. The maximum air pressure was 38 Ib per sq. in. In the case of sand face with poor leaky cover the usual practice was to make the air pressure equal to that of water from the surface down to about a quarter the distance below the top of the shield. The average amount of free air supplied per man per hour was approximately 2300 cub. ft. On the Hudson river side two tubes of the same size as those in the East River are for the Pennsylvania trains to New Jersey. Two tubes from Morton Street to New Jersey, begun by Haskin, already referred to, are for subway trains, and so are the most southerly of all on the Hudson side, viz. the two from Cortlandt Street to under the Pennsylvania station in Jersey City.
The two tubes from Morton Street were completed under the direction of Charles M. Jacobs, who was also chief engineer of the four other Hudson River tubes. The contractors for the Hudson tubes for the Pennsylvania road were the O'Rourke Contracting Company. Skilful treatment was required to overcome the difficulties on the New York side of the Hudson in all the tubes where the face excavation was partly in rock and partly in soft earth. Most of their length, however, was through silt, and in this the tunnelling was the easiest and most rapid that has ever been carried out in subaqueous work, 50 lineal ft. per day being sometimes accomplished. A large proportion of the silt which under ordinary processes would be taken into the tunnel through the shield, carried to the shore and got rid of by expensive methods, was by the latter process merely displaced as the shield with nearly or quite closed diaphragm was pushed ahead.
The East Boston tunnel, the first important example of a shield-built monolithic concrete arch, from the Boston Subway to East Boston, is 1-4 m. long, 3400 ft. being under the harbour. One mile was excavated by tunnelling with roof shields about 29 ft. wide, through clay containing pockets of sand and gravel. The engineer was H. A. Carson, and the contractors the Boston Tunnel Construction Company and Patrick McGovern.
Some 25 m. of waterworks brick-lined tunnels have been built since 1864, mostly in clay, under the Great Lakes, without the use of shields, though in the later ones compressed air was utilized. A large portion of the latest Cleveland tunnel, 9 ft. interior diameter, was built at the rate of 17 ft. per day at a cost of about $18 per ft. During this work three explosions of inflammable gases occurred, in which nineteen men were killed and others were injured. Later a fire at the shaft in the lake caused the death of ten men. Work was thereafter completed under the engineering direction of G. H. Benzenberg. Less serious accidents, principally explosions of marsh gas, occurred in many of the other tunnels. In one case (at Milwaukee under Benzenberg) drift material was penetrated, with large boulders and -coarse and fine gravel, and without any sand or clay filling, apparently in direct communication with the lake bottom. At times the necessary air pressure was 42 Ib per sq. in.
Subaqueous Tunnels made by sinking Tubes, Caissons, etc. In 1845 De la Haye, in England, doubtless having in mind the tedious and difficult work of the Thames tunnel, proposed to make tunnels un'der water by sinking large tubes on a previously prepared bed and connecting them together. Since then many inventors have proposed similar schemes. In 1866 Belgrand sank twin plate-iron pipes, i metre diameter and 156 metres long; under the Seine at Paris for a sewer siphon, and there have since been numerous examples of sunk cast-iron subaqueous water-pipes. It is believed that the first tunnel of this class, large enough for men to move upright in, was by H. A. Carson, assisted by W. Blanchard and F. D. Smith, in 1893-1894, in the outer portion of Boston harbour, for the metropolitan sewer outlet. The later tubes were about 9 ft. exterior diameter, in sections each 52 ft. long, weighing about 210,000 Ib, made of brick and concrete, with a skin of wood and water-tight bulkheads at each end. A trench was dredged in the harbour bed and saddles were accurately placed to support the tubes. The latter, made in cradles above water alongside a wharf, were lowered by long vertical screws moved by steam power, and were towed 5 to J m. to their final positions. After sufficient water had been admitted they were lowered to their saddles by travelling shears on temporary piles. The temporary joints between consecutive sections were made by rubber gaskets between flanges which were bolted together by divers. The later operations were backfilling the trench over the pipes, and in each section pumping out the water> removing its bulkheads, and making good the masonry between consecutive bulkheads, this masonry being inside the flanges. This work, about 1500 ft. in length, was done without contractors, by labourers and foremen under the immediate control of the engineers, and was found perfectly tight, straight and sound.
The double-track railroad tunnel at Detroit, made in 1906- 1909, under the direction of an advisory board consisting of W. J. Wilgus (chairman), H. A. Carson and W. S. Kinnear (the last-named being chief engineer), is 15 m. long, with a portion directly under the river of % m. The method used under the river (proposed by Wilgus) is an important variation on the Boston scheme. A trench was dredged with a depth equal to the thickness of the tunnel below the river bed and about 70 ft. below the river surface, and grillages were accurately placed in it to support the ends of thin steel tube-forms, inside of which concrete was to be moulded and outside of which deposited. These tubes, each about 23 ft. in diameter and 262-5 ft. long, were in pairs (one tube for each track), and were connected sidewise and surrounded by thin steel diaphragms 12 ft. apart. Planking, to limit the concrete, was secured outside the diaphragms (see fig. 3). The forms were made tight, bulkheaded at their ends, floated into place, sunk by admitting water, set on the grillages, and the ends of successive pairs connected together by bolts through rubber gaskets and flanges. The succeeding pair of tubes was not lowered until concrete had been deposited through the river around the tubes of the preceding pair. The following steps were to remove the water from one pair of tubes, mould inside a lining of concrete 20 in. thick, remove the contiguous bulkheads, and repeat again and again the processes described until the subaqueous tunnel was complete.
The New York Rapid Transit tunnel under Harlem river, built 1904-1905, has two tubes, each about 15 ft. diameter and 400 ft. long, with a surrounding shell of cast iron itself surrounded by concrete. The outside width of concrete is about 33 ft. Its top is 28 ft. below high water and about 3 ft. below the bed of the river. D. D. McBean, the sub-contractor, dredged a trench in the river to within 7 or 8 ft. of the required depth. He then enclosed a space of the width of the tunnel from shore to mid-stream with i2-in. sheet piling, which was evenly cut off some 2 ft. above the determined outside top of the tunnel. On top of this piling he sank and tightly fitted a flat temporary roof of timber 3 ft. thick in sections, and covered this with about 5 ft. of dredged mud. Water was expelled from this subaqueous chamber by compressed air, after which the remaining earth was easily taken out, and the iron and concrete tunnel walls were then built in the chamber. For the remaining part of the river the foregoing process was varfed by cutting off the sheet piling at mid-height of the tunnel and making the upper half of the tunnel, which was built above and lowered in sections through the water, serve as the roof of the chamber in which the lower half of the tunnel was built.
The tunnels of the Metropolitain railway of Paris (F. Bienventie, engineer-in-chief) under the two arms of the Seine, between Place Chatelet and Place Saint Michel, were made by means of compressed-air caissons sunk beneath the river bed, were next made by the aid of temporary small caissons sunk through about 26 ft. of earth under the river. The tops of the side walls were made even with the end walls. A steel rectangular coffer-dam (figs. 5 and 6) was sunk to rest with rubber or clay joint on these surrounding walls. The coffer-dam had shafts reaching above the surface of the water, so that the earth core was easily taken out (after removing the water) in free air. The adjacent chambers under the caissons were then connected together. Three caissons, of a total length of 396 ft., were used under the larger arm, and two, of an aggregate length Mountain Tunnels for Railways.
Internal Width and Height.
Average progress per day = 24 hrs. (lin. yds.).
Approximate cost per lin. yd.
Mont Cenis (l tunnel) . St Gotthard (i tunnel) Arlberg (i tunnel) .
7-98 9'3 6-36 12-3 26 ft. 3 in. X 24 ft. 7 in. (horseshoe). 26 ft. 3 in. X 24 ft. 7 in. (horseshoe). 25 ft. 3 in. wide 16 ft. 5 in. X 19 ft. 6 in. each (rain.).
Granitic Granitic Gneiss, mica schist, limestone and disintegrated mica schist rock.
2'57 6-61 9-07 11-63 L 226 143 108 148 L. Chagnaud being the contractor. They were built of plates of sheet steel and masonry, with temporary steel diaphragms in the ends, filled with concrete, making a cross wall with a level top about even with the outside top of the tunnel and about 2 ft. below the bottom of the Seine. The caissons were sunk on the line of the tunnel so that adjacent ends (and the walls just described) were nearly 5 ft. apart with at that stage a core of earth between them. Side walls joining the end walls and thus enclosing the earth core on four sides (fig. 4)
Concrete Wbfls on %p of Ca/sson Ends^f- / MMM *l4 < -Lateral Walls J- -r- sef with Temporary Cff/ssot >'' ""k'1 (From Engineering Neva, New York.)
FIG. 4. Perspective showing manner of enclosing space between tunnel caissons for the Metropolitain under the Seine at Paris.
(From Engineering News, New York.)
FIG. 5. Transverse Section.
Coffer-dam superimposed over joints between caissons in tunnels for the Metropolitain under the Seine at Paris.
of 132 ft., under the smaller arm of the Seine. The cost of the tunnel was 7000 francs per lineal metre.
William Sooy Smith published in Chicago, in 1877, a description of a scheme for building a tunnel under the Detroit river by sinking caissons end to end, each caisson to be secured to the adjoining one by tongued and grooved guides, and a nearly water-tight connexion between the two to be made by means of an annular inflated hose.
Tunnelling through Mountains. Where a great thickness of rock overlies a tunnel through a mountain, it may be necessary to do the work wholly from the two ends without intermediate shafts. The problem largely resolves itself into devising the most expeditious way of excavating and removing the rock. Experience has led to great advances in speed and economy, as may be seen from examples in the above table.
In 1857 the first blast was fired in connexion with the Mont Cenis works; in 1861 machine drilling was introduced; and in 1871 the tunnel was opened for traffic. With the exception of about 300 yds. the tunnel is lined throughout with brick or stone. During the first four years of hand labour the average progress was not more than 9 in. per day on each side of the Alps; but with compressed air rock-drills the rate towards the end was five times greater.
In 1872 the St Gotthard tunnel was begun, and in 1 88 1 the first locomotive ran through it. Mechanical drills were used from the beginning. Tunnelling was carried on by driving in advance a top heading about 8 ft. square, then enlarging this sideways, and finally sinking the excavation to invert level (see figs. 7 and 8). Air for working the rock-drills was compressed to seven atmospheres by turbines of about 2000 horse-power.
The driving of the Arlberg tunnel was begun in 1 880 and the work was completed in little more than three years. Themainheading was driven along the bottom of the FIG. 6. Longitudinal Section.
tunnel and shafts were opened up 25 to 70 yds. apart, from which smaller 'headings were driven right and left. The tunnel was enlarged to its full section at different points simultaneously in lengths of 8 yds., the excavation of each occupying about twenty days, and the masonry fourteen days. Ferroux percussion air-drills and Brandt rotary hydraulic drills were used, the performance of the latter being especially satisfactory. After each blast a fine spray of water was injected, which assisted the ventilation FIGS. 7 and 8. Method of excavation in St Gotthard Tunnel.
materially. In the St Gotthard tunnel the discharge of the air-drills was relied on for ventilation. In the Arlberg tunnel over 8000 cub. ft. of air per minute were thrown in by ventilators. To keep pace with the miners, 900 tons of excavated material had to be removed, and 350 tons of masonry introduced, daily at each end of the tunnel, which necessitated the transit of 450 wagons. The cost per lineal yard varied according to the thickness of masonry lining and the distance from the mouth of the tunnel. For the first thousand yards from the entrance the prices per lineal yard were 11 8s. for the lower heading; 7 123. for the upper one; 30 los. for the unlined tunnel; 45 for the tunnel with a thin lining of masonry; and 124 55. with a lining 3 ft. thick at the arch, 4 ft. at the sides, and 2 ft. 8 in. at the invert.
The Simplon tunnel was begun in 1898 and completed in 1905. It is over 30 % longer than the St Gotthard, and the greatest depth below the surface is 7005 ft. A novel method was introduced in the shape of two parallel bores (56 ft. apart, connected at intervals of 660 ft. by oblique galleries), which greatly facilitated ventilation, and resulted in increased economy and rapidity of construction, while ensuring the health of the men. One of these galleries was made large enough for a singletrack railroad, and the second is to be enlarged and similarly used. The death-rate in the Simplon tunnel was decreased as compared with the St Gotthard from 800 in eight years to 60 in seven years. Had one wide tunnel been made instead of two narrow ones, it would have been difficult to maintain its integrity; even with the narrow cross-section employed the floor was forced up at points in the solid rock from the great weight above, and had to be secured by building heavy inverts of masonry. Temperatures were reduced to 89 F. by spraying devices, although the rock temperatures ranged from 129 to 130 F. At one point 4374 yds. from the portal of Iselle the " Great Spring " of cold water was struck; it yielded 10,564 gallons per minute at 600 ft pressure per sq. in., and reduced the temperature to 55-4 F., the lowest point recorded. A spring of hot water was met on the Italian side which discharged into the tunnel 1600 gallons per minute with a temperature of 113 F. The maximum flow of cold water was 17,081 gallons per minute, and of hot water 4330 gallons per minute. These springs often necessitated a temporary abandonment of the work. Water power from the Rhone at the Swiss and from the Diveria at the Italian end provided the power for operating all plant during the construction of most of the work. Among the able engineers connected with this work must be mentioned Alfred Brandt, a man of remarkable energy and ability, whose drills were used with much success. He died early in the work, of injuries received from falling rock.
A group of tunnels the Tauern, Barengraben, Wocheiner and Bosriick was undertaken by the Austrian government in connexion with new Alpine railroads to increase the commercial territory tributary to the seaport of Trieste, which at one time was greater than Hamburg. The principal tunnel of this group is under the main body of the Tauern mountain. The bottom drifts met on the 21st of July 1907. The difficulties resulted mostly from mountain debris and springs. There are four minor tunnels between Schwarzach, St Veit, and the north portal of the Tauern, and nineteen between the south portal and the south slope at Mollbriicken.
The electric railway from the Eiger glacier to near the summit of the Jungfrau includes a tunnel 15 m. long, 3-6 metres wide and 3-8 metres high, with a midway station, from which a large part of northern Switzerland can be seen. From the Jungfrau terminus, at an elevation of 13,428 ft., the summit, 242 ft. higher, will be reached by an elevator.
The Hoosac tunnel was the first prominent tunnel in America. It was begun in 1855 and finished in 1876, after many interruptions. It was memorable for the original use in America of air-drills and nitroglycerin. The Pennsylvania railroad tunnels crossing New York City under 32nd and 33rd Streets are of unusual size. Owing to the close proximity of large buildings and other structures special methods were adopted for mining the rock to lessen the vibrations by explosions. At 33rd Street and 4th Avenue the tunnels pass directly under two of the Rapid Transit system, above which there is another belonging to the Metropolitan Traction Company, so that there are three tunnels at different levels under the street.
Among other rock tunnels may be mentioned the Albula, through a granite ridge of the Rhaetian Alps, for a single-track narrow-gauge railroad, 3-6 m. long; tunnels on the Midland railway, near Totley in Derbyshire, over 3-5 m. long, largely in shale, and at Cowburn, over 2 m. long, in shale and harder rock, each 27 ft. wide and 20-5 ft. high inside; the Suram, on the Trans-Caucasus railway, for double track, 2-47 m. long, through soft rock; the tail-race tunnel for the Niagara Falls Water Power Company, 1-3 m. long, 19 ft. wide and 21 ft. high, through argillaceous shale and limestone, costing about $1,250,000; the Tequixquiac outlet to the drainage system for the city of Mexico, costing $6,760,000; the Cascade, Washington, part of the Great Northern railroad system, saving 9 m. in distance; and the Gunnison, irrigating 147,000 acres in Colorado.
Tunnelling in Towns. Where tunnels have to be carried through soft soil in proximity to valuable buildings special precautions have to be taken to avoid settlement. A successful example of such work is the tunnel driven in 1886 for the Great Northern Railway Company under the Metropolitan Cattle FIG. 9. Paris Metropolitain Tunnel, longitudinal horizontal section.
Market, London. This was done by the crown-bar method, the bars being built in with solid brickwork. The subsidence in the ground was from i to about 3^ in. Several buildings were tunnelled under without any structural damage.
London has now some 90 m. of tunnels for railways, mostly operated by electric traction. Most of those which have been constructed since 1890 have been tunnelled by the use of cylindrical shields and walls of cast iron. Shields about 23 ft. in diameter were used in constructing the stations on the Central London railway, and one 32 ft. 4 in. in diameter and only 9 ft. 3 in. long was used for a short distance on the Clapham extension of the City and South London railway.
general, the upper half of the tunnel was executed first (figs. 9 and 10) and the lower part completed by underpinning.
Figs, n, 12 and 13 illustrate a case of tunnelling near important buildings in Boston in 1896, with a roof-shield 29 ft. 4 in. in external diameter. The vertical sidewalls were first made in small drifts, the roof-shield running on top of these, and the core was taken out later and the invert or floor of the tunnel put in last. Each hydraulic press of the shield reacted against a small continuous cast-iron rod imbedded in the brick arch. In some large sewerage tunnels in Chicago the shields were pushed from a wall of oak planks, 8 in. thick, surrounding the brick walls of the sewer.
FIG. 10. Paris Metropolitain Tunnel, longitudinal vertical section.
Paris has an elaborate plan for underground railways some 50 m. in length, a considerable number of which have been constructed since 1898 under the engineering direction of F. Bienveniie. Instead of using completely cylindrical shields and cast-iron walls, as in London, roof-shields (boucliers de votile) were employed for the construction of the upper half of the tunnel, and masonry walls were adopted throughout. In Ventilation oj Tunnels. The simplest method for ventilating a railway tunnel is to have numerous wide openings to daylight at frequent intervals. If these are the full width of the tunnel, at least 20 ft. in length, and not farther apart than 200 yds., it can be naturally ventilated. Such arrangements are, however, frequently impracticable, and then recourse must be had to mechanical means.
FIG. ii. Boston Subway, first and second phases.
FIG. 12. Boston Subway, third phase.
FIG. 13. Boston Subway, longitudinal vertical section through shield.
The first application of mechanical or fan ventilation to railway tunnels was made in the Lime Street tunnel of the London and North-Western railway at Liverpool, which has since been replaced by an open cutting. At a later date fans were applied to the Severn and Mersey tunnels.
The principle ordinarily acted upon, where mechanical ventilation has been adopted, is to exhaust the vitiated air at a point midway between the portals of a tunnel, by means of a shaft with which is connected a ventilating fan of suitable power and dimensions. In the case of the tunnel under the river Mersey (fig. 14) such a shaft could not be provided, owing to the river being overhead, but a ventilating heading was driven from the middle of the river (at which point entry into the tunnel was effected) to each shore, where a fan 40 ft. in diameter was placed. In this way the vitiated air is drawn from the lowest point of the railway, while fresh air flows in at the stations on each side to replenish the partial vacuum, as indicated by arrows in the accompanying longitudinal section of the tunnel. The principle was that fresh air should enter at each station and " split " each way into the tunnel, and that thus the atmosphere on the station platforms should be maintained in a condition of purity.
The fans in the Mersey tunnel are somewhat similar to the wellknown Guibal fans, with the exception of an important alteration in the shutter. With the Guibal shutter, the top of the opening (From a diagram in Proc. Inst. Civ. Eng.)
Longitudinal Section of the Mersey Tunnel, showing Method of Ventilation.
into the chimney from the fan has a line parallel to that of the fanshaft and of the fan-blades, and, as a consequence, as each blade passes this shutter, the stoppage of the discharge of the air is instantaneous, and the sudden change of the pressure of the air on the face cf the blade whilst discharging and the reversal of the pressure, due to the vacuum inside the fan-casing, cause the vibration hitherto inseparable from this type of ventilator. As an illustration of the effect of the pulsatory action of the Guibal shutters the following figures may be given : a fan having ten arms and running, say, sixty revolutions per minute, and working twenty-four hours per day, gives (10 X 60 X 60 X 24 = ) 864,000 blows per day transmitted From the tip of the fan-vanes to the fan-shaft; the shaft is thus in a constant state of tremor, and sooner or later reaches its elastic limit, and the consequent injury to the general structure of the fan is obvious. This difficulty is avoided by cutting a A -shaped opening in the shutter, thus gradually decreasing the aperture and allowing the air to pass into the chimney in a continuous stream instead of intermittently. The action of this regulating shutter increases the durability and efficiency of the fans in an important degree. In towns like Liverpool and Birkenhead any pulsatory action would be readily felt by the inhabitants, but with the above arrangement it is difficult to detect any sound whatever, even when standing close to the buildings containing the fans. The admission of the air on both sides is found in practice to conduce to smooth running and to the reduction of the side-thrust which occurs when the air is admitted on one side only. The fans are five in number: two are 40 ft. in diameter by 12 ft. wide, and two 30 ft. in diameter by loft, wide, one of each size being erected at Liverpool and at Birkenhead respectively. In addition, there is a high-speed fan 1 6 ft. in diameter in Liverpool which throws 300,000 cub. ft.
The following table gives the result of experiments made with the ventilating fans of the Mersey railway :
a S l Fan at *o S "oil jj'*' o ^J BJ ft >% S3 *J rj S i' s 1zz .1 li 8 < > Hamilton Street, Birkenhead Shore Road, "3 1-30 1895 214,135 Birkenhead James Street, 1zz 1 2-50 3288' 134-685 Liverpool . James Street, 1zz 2 2-45 2465 178,880 Liverpool .
60 2-30 2062 123,720 Bold Street, Liverpool ..
1 6* 1zz 00,000 Total 951,420 The central point of the Severn tunnel (fig. 15) lies toward the Monmouthshire bank of the river, and ventilation is effected from that point by means of one fan placed on the surface at Sudbrooke, Monmouth, at the top of a shaft which is connected with a horizontal Ventilating Fan Monmouthshire 40 "^ f"''"V^L? el/em 1 Gloucestershire 1zz milei al length of Tunnel 4 miles 624 yardt rd FIG. 15. Section of Severn Tunnel (Fox).
heading leading to the centre. This fan, which is 40 ft. in diameter by 12 ft. in width, removes from the tunnel some 400,000 cub. ft. per minute, and draws in an equivalent volume of fresh air from the two ends.
About 1896 an excellent system was introduced by Signor Saccardo, the well-known Italian engineer, which to a great extent has minimized the difficulty of ventilating long tunnels under mountain-ranges where shafts are not available. This system, which is not applicable to tunnels in which underground stations exist, is illustrated in fig. 1 6, which represents its application to the single-line tunnel through the Apennines at Pracchia. This tunnel is one of fiftytwo single-line tunnels, with a gradient of I in 40, on the main line between Florence and Bologna, built by Thomas Brassey. There is a great deal of traffic which has to be worked by heavy locomotives. Before the installation of a ventilating system under any condition of wind the state of this tunnel, about 3000 yds. in length, was bad ; _* In the case of this circular drift-way a velocity of 4000 ft. per minute was subsequently attained. * Quick-running fan.
but when the wind was blowing in at the lower end at the same time that a heavy goods or passenger train was ascending the gradient the condition of affairs became almost insupportable. The engines, working with the regulators full open, often emitted large quantities of both smoke and steam, which travelled concurrently with the train. The goods trains had two engines, one in front and another at the rear, and when, from the humidity in the tunnel, due to the (From the Proc. Inst. Civ. Ens.)
FIG. 16. Diagram illustrating the Saccardo System for Ventilating Tunnels.
steam, the wheels slipped and possibly the train stopped, the state of the air was indescribable. A heavy train with two engines, conveying a royal party and their suite, arrived on one occasion at the upper exit of the tunnel with both enginemen and both firemen insensible; and on another occasion, when a heavy passenger train came to a stop in the tunnel, all the occupants were seriously affected.
In applying the Saccardo system, the tunnel was extended for 15 or 20 ft. by a structure either of timber or brickwork, the inside line of which represented the line of maximum construction, and this was allowed to project for about 3 ft. into the tunnel. The space between this line and the exterior constituted the chamber into which air was blown by means of a fan. Considering the length of tunnel it might at first be thought there would be some tendency for the air to return through the open mouth, but nothing of the kind happened. The whole of the air blown by the fan, 164,000 cub. ft. per minute, was augmented by the induced current yielding 46,000 cub. ft. per minute, making a total of 210,000 cub. ft.; and this volume was blown down the gradient against the ascending train, so as to free the driver and men in charge of the train from the products of combustion at the earliest possible moment. Prior to the installation of this system the drivers and firemen had to be clothed in thick woollen garments, pulled on over their ordinary clothes, and wrapped round and round the neck and over the head ; but in spite of all these precautions they sometimes arrived at the upper end of the tunnel in a state of insensibility. The fan, however, immensely improved the condition of the air, which is now pure and fresh.
In the case of the St Gotthard tunnel, which is c>$ m. in length and 26 ft. wide with a sectional area of 603 sq. ft., the Saccardo system was installed in 1899 with most beneficial results. The railway is double-tracked and worked by steam locomotives, the cars being lighted by gas. The ventilating plant is situated at Goschenen at the north end of the tunnel and consists of two large fans operated by water power. The quantity of air passed into the narrow mouth of the tunnel is 413,000 cub. ft. per minute at a velocity of 686 ft., this velocity being much reduced as the full section of the tunnel is reached. A sample of the air taken from a carriage contained 10-19 parts of carbonic acid gas per 10,000 volumes.
In the Simplon tunnel, where electricity is the motive power, mechanical ventilation is installed. A steel sliding door is arranged at each entrance to be raised and lowered by electric power. After the entrance of a train the door is lowered and fresh air forced into the tunnel at considerable pressure from the same end by fans.
The introduction of electric traction has simplified the problem of ventilating intra-urban railways laid in tunnels at a greater or less distance below the surface, since the absence of smoke and products of combustion from coal and coke renders necessary only such a quantity of air as is required by the passengers and staff. For supplying air to the shallow tunnels which form the underground portions of the Metropolitan and District railways in London, open staircases, blow-holes and sections of uncovered track are relied on. When the lines were worked by steam locomotives they afforded notorious examples of bad ventilation, the proportion of carbonic acid amounting to from 15 or 20 to 60, 70 and even 89 parts in 10,000. But since the adoption of electricity as the motive power the atmosphere of the tunnels has much improved, and two samples taken from the cars in 1905 gave 11-27 an d 14-07 parts of carbonic acid in 10,000.
When deep level " tube " railways were first constructed in London, it was supposed that adequate ventilation would be obtained through the lift-shafts and staircases at the stations, with the aid of the scouring action of the trains which, being of nearly the same cross-section as the tunnel, would, it was supposed, drive the air in front of them out by the openings at the stations they were approaching, while drawing fresh air in behind them at the stations they had left. This expectation, however, was disappointed, and it was found necessary to employ mechanical means. On the Central London railway, which runs from the Bank of England to Shepherd's Bush.adistance of 6 m.,the ventilating plant installed-in 1902 consists of a 300 h.p. electrically driven fan, which is placed at Shepherd's Bush and draws in fresh air from the Bank end of the line and at other intermediate points. The fan is 5 ft. wide and 20 ft. in diameter, and makes 145 revolutions a minute, its capacity being 100,000 cub. ft. a minute. It is operated from I to 4 a.m., and the openings at all the intermediate stations being closed it draws fresh air in at the Bank station. The tunnel is thus cleared out about 2 1 times each night and the air is left in the same condition as it is outside. The fan is also worked during the day from n a.m. to 5 p.m., the intermediate doors being open; in this way the atmosphere is improved for about half the length of the line and the cars are cleared out as they arrive at Shepherd's Bush. Samples of the air in the tunnel taken when the fan was not running contained 7-07 parts of carbonic acid in 10,000, while the air of a full car contained 10-7 parts. The outside air at the same time contained 44 parts. A series of tests made for the London County Council in 1902 showed that the air of the cars contained a minimum of 9-60 parts and a maximum of 14-7 parts. In some of the later tube railways in London such as the Baker Street and Waterloo, and the Charing Cross and Hampstead lines electrically driven exhaust fans are provided at about half-mile intervals; these each extract 18,500 cub. ft. of air per minute from the tunnels, and discharge it from the tops of the station roofs, fresh air being conveyed to the points of suction in the tunnels.
Tne Boston system of electrically operated subways and tunnels is ventilated by electric fans capable of completely changing the air in each section about every fifteen minutes. Air admitted at portals and stations is withdrawn midway between stations. In the case of the East Boston tunnel, the air leaving the tunnel under the middle of the harbour is carried to the shore through longitudinal ducts (fig- 3) atl d is there expelled through fan-chambers.
In the southerly 5 m. of the New York Rapid Transit railway, which runs in a four-track tunnel of rectangular section, having an area of 650 sq. ft., and built as close as possible to the surface of the streets, ventilation by natural means through the open staircases at the stations is mainly relied upon, with satisfactory results as regards the proportions of carbonic acid found in the air. But when intensely hot weather prevails in New York the tunnel air is sometimes 5 hotter still, due to the conversion of electrical energy into heat. This condition is aggravated by the fine diffusion through the air of oil from the motors, dust from the ballast and particles of metal ground off by the brake shoes, etc.
Volume of Air Required for Ventilation. The consumption of coal by a locomotive during the passage through a tunnel having been ascertained, and 29 cub. ft. of poisonous gas being allowed for each pound of coal consumed, the volume of fresh air required to maintain the atmosphere of the tunnel at a standard of purity of 20 parts of carbon dioxide in 10,000 parts of air is ascertained as follows: The number of pounds of fuel consumed per mile, multiplied by 29, multiplied by 500, and divided by the interval in minutes between the trains, will give the volume of air in cubic feet which must be introduced into the tunnel per minute. As an illustration, assume that the tunnel is a mile in length, that the consumption of fuel is 32 ft per mile, and that one train passes through the tunnel every five minutes in each direction; then the volume of air required per minute will be 32 ft X 29 cub, ft. X 500 = cub f 2j minutes.
Corrosion of Rails in Tunnels. Careful tests made in the Box and Severn tunnels of the Great Western railway, to ascertain if possible the loss that takes place in the weight of rails owing to the presence of corrosive gases, gave the following results:
Box TUNNEL (i m. 66 chains in length).
Percentage of Wear per annum.
ft per yard Down line, gradient falling I in 100 % per annum.
At east mouth 0-439 = 0-377 28 chains from east mouth I 800 = I -540 48 chains from east mouth 2-110 = 1-810 I m. 8 chains from east mouth 2-880 = 2.480 At west mouth 0-640 = 0-553 Up line, gradient rising I in 100 At east mouth 0-620 = 0-575 I m. 8 chains from east mouth i -500 = I -380 i m. 28 chains from east mouth 1520 = 1-310 At west mouth 0-680 = 0-587 SEVERN TUNNEL (4 m. 28i chains in length). Percentage of Wear per annum.
Ib per yard JJown line, outside and quite clear of tunnel, % per annum.
Bristol end, gradient falling I in 100 . . . 0-280 = 0-240 Up line, outside and quite clear of tunnel, Newport end, gradient falling I in 90 . . 0-440 = 0-390 At Bristol mouth, gradient falling I in 100 1-200 = 1-020 33 chains from Bristol mouth, gradient falling 1 in iop 2-160 = 1 860 3 m - 75i chains from Bristol mouth, gradient rising I in 90 1-900=1-630 At Newport mouth 0-310 = 0-270 Down and up line under main-shaft level . . 3-200 = 2-750 It will be seen that the maximum wear and corrosion together reached the extraordinary weight of 2 \ ft per yard of rail per year a very serious amount that involved great expenditure The wear occurred over the whole of the rail, but the top, over which the engine and train passed, wore at a greater rate, presumably on account of the surface being kept bright and the gases being able to act on it. The Great Western Company tried the experiment in the Severn tunnel of boxing up the rails, so that the ballast approached their surface within i in. or i^ in. It was found, however, that in the case, at any rate, of the limestone ballast the cure was almost worse than the disease, the result being a maximum wear of 2j ft and an average wear of just under 2 ft> per yard of rail per year. The average on the open line would be about 0-25 ft in the same time.
See Proc. Inst. Civ. Eng. ; also works on tunnelling by Drinker, Simms, Stauffer and Prelini, and on tunnel shields, etc., by Copperthwaite. (H. A. C.)
Note - this article incorporates content from Encyclopaedia Britannica, Eleventh Edition, (1910-1911)