ÆäÀÌÁö À̹ÌÁö
PDF
ePub

Probably the imperfection of the old | allowed a greater loss in its few inches of cable was due rather to the joints between length than occurred from some miles of the separate miles of wire as manufactured, sound cable. A bad joint seldom does more than to any extreme inferiority in the gutta- than this at first, but in time it becomes percha employed. These joints are even brittle, cracks, leaves the sound gutta-percha now the weak places in the protection of a at each side, and, finally, allows the water cable. When the gutta-percha has been free access to the strand. Joints of this selected and purified with care, and applied character have been found in considerable by mechanical contrivances of proved ex- number in old cables, and especially in the cellence, there is little risk of a fault occur- old 1857-58 Atlantic Cable. Some of these ring; but this manufacture cannot be so present an appearance of extraordinary conducted as to produce one unbroken length carelessness, even the copper strands being of wire, and even if it could, convenience in imperfectly joined. It is almost certain the other processes of manufacture would that the final failure of the 1858 Atlantic require the division of this wire into lengths. Cable was due to one of these joints in which One-mile lengths are, in practice, usually the copper was imperfectly joined; the made without joint, and are joined together wires were pulled asunder when the cable by a skilled workman as occasion arises. was being laid, they came together again The copper strands are soldered together when the strain was removed, but the points with a scarf-joint, two pieces of fine wire are of contact soon were oxidized, and all comthen wrapped over this joint, so that even munication ceased. Mere loss of insulation if it is pulled asunder, electrical continuity hardly ever entirely stops signals. will be preserved, and so far the operation is one of no great difficulty. This cannot be said of the next process, the insulation of the wire by hand, and the welding, as it were, of the new sheets of gutta-percha, so applied with the old sheathing on either side. The gutta-percha is warmed by a spirit-lamp; too much or too little heat is fatal, and the jointer must judge of the temperature by experience; the least moisture will spoil a joint,-hence one reason for providing that no moisture can percolate along the metal strand. A very little dirt or impurity will also do much injury, hence the rule that a jointer must do no other work, and that the copper wire must be soldered by one man, the gutta-percha applied by another. A joint may also be spoilt by the presence of air under one of the insulating coats, and as the writer cannot pretend himself to make a joint, other causes of failure probably exist of which he is ignorant, but enough has been said to show the difficulty of the process. Fortunately, joints can now be tested apart from the rest of the cable. In old times when a joint had been made the whole cable was tested; if the leak from the new joint was inconsiderable in comparison with the loss from the whole cable, perhaps some hundred miles long, the joint was supposed to be good, although, perhaps, it may have

[blocks in formation]

The test now employed shows whether a joint is as good as any equal length of the wire, and all joints which do not reach this standard are mercilessly cut out. First the joints to be tested are allowed to soak in water for twenty-four hours, then they are placed in an insulated trough of water connected with a Leyden jar of large surface, the cable is charged with a powerful battery, and a little electricity leaks out through the joints into the insulated trough. If the joint is good, this leakage is so small that the current produced by it could not be shown by the most sensitive galvanometer, but after a minute or two minutes, the insulated trough and Leyden jar will be charged by the gradual accumulation of electricity which has slowly leaked through the joint. If this be now discharged through a galvanometer, it will produce a sensible effect, and can be measured. In fact, the leak which was too small to be directly perceptible, is not only perceived, but its amount ascertained by measuring the quantity which accumulates from it in a given time. This test is due to Messrs. Bright and Clark. Other tests of a similar nature have been proposed, but have been found less convenient. The first test for a joint, distinct from that of the whole cable, was, it is believed, proposed by Mr. Whitehouse. No instance has yet occurred of failure in a joint which has successfully passed the accumulation tests above described. There are about two thousand joints in each Atlantic Cable.

Any further description of the various tests would only be wearisome. There are tests of charge, of discharge, of the effects of electrification, of the effects of positive

and negative currents, tests with statical electricity as well as voltaic currents; but enough has been said to show that the examination of a submarine cable, as now conducted is not guess-work, or even a matter of experience and skill; it consists simply of a long and laborious series of exact measurements, so expressed in figures that all electricians can understand the results, and compare them with those obtained from other cables, or by other observers. In this lies our safety. Granting that the production of a perfectly insulated conductor 2000 miles long is no longer a matter of chance, can we protect and lay this wire with equal certainty in such depths as the Atlantic presents? or do we here fall back into a region of mere good or bad luck? As to shallow water, the question need not be asked. No serious strains occur, and the submersion of the cable depends on a few simple mechanical arrangements which have long since been perfected. Even in deep water, cables have not broken during the laying nearly so often as is supposed. Some very early Mediterranean expeditions, a later attempt to join Candia with Alexandria, and the experimental trip of the first Atlantic expedition, give almost the only instances where a cable parted suddenly during submersion; but it must be allowed that the strains endured in passing over depths of 2000 fathoms approached far too nearly to the breaking strain of the cables, and it is by no means impossible that some cables may have been injuriously stretched, although they were not broken.

In order to lay a cable of any construction taut along the bottom of the sea, it is necessary to restrain its free exit from the ship by applying a retarding force nearly equal to the weight of a length of the cable, hanging vertically from the ship to the bottom of the sea. Cables of the old form, in which simple iron wires were laid round its core, would support from 4000 to 5000 fathoms of themselves hanging vertically in water. They could, therefore, be laid fairly taut in depths of 2000 or 2500 fathoms, such as are met with in the Atlantic, but engineers are in the habit of allowing a very much larger margin than the above. They make all their structures from six to ten times stronger than by exact calculation they need be. This figure 'six' or 'ten' they call the co-efficient of safety. A co efficient of safety of 'two,' such as was given by these old cables, gave very little safety indeed. When the cables are not laid taut, but with a certain slack, the strain need not be quite so great. The friction of the water tends to relieve the strain, but this relief with the old smooth cables was small.

Sir W. Thomson was again the first to give the true theory of the strains which occur, and the curve assumed by the rope during submersion. The first account of the theory appears in the Engineer newspaper of October 1857.

A much more elaborate investigation was, independently of Sir W. Thomson's theory, made by Messrs. Brook and Longridge, whose able paper was published in the Transactions of the Institution of Civil Engineers for 1858. Dr. Siemens of Berlin independently arrived at similar conclusions; the subject is nevertheless not a very simple one, for the Astronomer-Royal was misled more than once in his investigations concerning it.

When the ship and cable are both at rest, the latter hangs in a simple catenary curve, the strains on which are easily computed; but when the cable is being payed out, it lies in an inclined straight line from a point a very little below the surface of the sea to the bottom (provided, however, the cable as it lies at the bottom is not strained); above the water the cable hangs in a short catenary; the angle at which the cable lies in the water depends on the speed of the ship, and the specific gravity of the cable; it is independent of the strain on the cable, and is therefore unaltered whether the cable is being payed out slack or taut. As the speed of the ship increases, the angle which the cable makes with the horizon diminishes; the same effect is produced by diminishing the specific gravity of the cable-that is to say, by increasing its bulk relatively to its weight. The Atlantic Cable, under the water, probably lay at an angle of nearly 7° with the horizon; on leaving the ship, the angle was 9. In this case, in a depth of two miles, a length of from 16 miles cable would lie in the water between the point where it left the ship and that where it touched the bottom. The weight of this cable, weighed in water, would be 231 cwt.; fortunately, as the cable would break with about 153 cwt., only a very small part of this weight is borne by the cable itself as it leaves the ship. Even if the cable were to be laid absolutely taut, a restraining force of 28 cwt. only would be necessary. In practice, 12 cwt. to 14 cwt. was found quite sufficient.

The cable, as it leaves the ship, may almost be said to lie on a long inclined plane of water; if it lay on a solid inclined plane, without friction, it might, by a well-known law of mechanics, be balanced by a length of itself hanging vertically from the apex of the inclined plane to the bottom, and this is almost exactly the strain required to be given by the break on board ship to balance the

250

Submarine Telegraphy.

cable, or, in other words, to prevent it from shooting back along the inclined plane, so as to lie slack in folds at the bottom; but the inclined plane of water is not at rest, it yields under the cable at every instant, at every spot; yet if the cable were pressed through the water, so that the water yielded before it, but did not slip along it at all, the analogy of the inclined plane would be quite perfect. The resistance of the water to displacement would supply the component of the whole force required, perpendicular to the direction of the cable, exactly as in the case of a solid plane; but, on constructing a diagram, it will at once be seen that the cable, as it descends, slips a little along the plane, and the friction of the water opposing this slip, slightly diminishes the strain required to lay the cable taut. If, on board ship, this full strain is not produced by the breaks, the cable slips still faster back along the inclined plane, and with such a velocity that the friction of the water on the cable makes up for the insufficient tension given by the breaks, and equilibrium is again restored, but at the expense of a waste of cable. It will be clear that, with a given depth, the greater the length of cable in the water, the less need this waste be, for the friction will be directly proportional to the surface; further, for the same reason, the waste will be less the more bulky the cable, and the rougher the surface. iron cables of small diameter and smooth With the old surface, very little advantage was gained by diminishing the strain on the breaks below that due to the full depth of water; a very slight relief of strain was followed by a perfect rush of cable out of the ship, and a loss of twenty or twenty-five per cent. was followed by a comparatively small diminution in the risk of fracture. In the cables of the Atlantic class, the bulk relatively to the weight is very greatly increased by enveloping each iron or steel wire in a separate covering of hemp, before laying them round the gutta-percha. These cables lie at a much smaller angle with the horizon, they offer a much larger and rougher surface than the simple iron cable, and consequently the friction, as they run back on the inclined waterplane, is very much larger. With cables of that class it becomes practicable and desirable to diminish the strain produced by the break much below that due to the full depth of water. Slack to the amount of twelve or fifteen per cent. diminishes the necessary strain on the breaks by more than onehalf, and the importance of this relief can hardly be over-estimated. It actually becomes practicable to disregard the depth over which the ship is passing. The breaks

Dec.

may be set to give the strain thought desirable, and the cable will then take care of itself. In shallower water, less slack will be payed out, in deeper water more, but the amount is never excessive, and can at any time be diminished by increasing the speed of the ship, which, by diminishing the angle at which the cable lies with the horizon, augments the effect of the friction of the inclined water-plane. This effect must not be confounded with the effect that would be produced by a buoyant substance attached to the cable. The hemp is no lighter than water, and does not tend by its buoyancy to carry any part of the weight of the cable, but it increases the bulk, and therefore increases the resistance of the water to displacement, and both directly and indirectly increases the surface friction.

during submersion was from 12 to 14 cwt.; The strain on the new Atlantic Cables their strength is 150 or 160 cwt. there is a co-efficient of safety of ten instead of two or four. The first cable out of Here the water weighed little more than half as much as the new cables; in water, it weighed more than they do. Its strength was 80 cwt., and the maximum strain during its submersion was nearly one ton; the ordinary strains varied from 1500 to 1900 lbs.

ress which has been made in the mechanical From the figures, we may learn the progished risk which attends their submersion. construction of the cables, and the dimin

Before

the cables helps to show the progress made The history of the several attempts to lay in the construction, and bears out the conclusions as to the improvements effected. In August 1857 a first attempt was made to lay an Atlantic Cable; 330 knots were laid, starting from Valentia. Then the cable broke, the indicated strain being about 27 cwt. The retarding friction on this occasion was produced by two blocks of wood which the next attempt the Appold break had been were clamped round a small drum. invented, and with the sanction of Mr. Penn, Mr. Field, Messrs. Easton and Amos, Mr. Lloyd, Mr. Everett, and Sir C. Bright, it was applied to the paying-out machinery. This break is an excellent contrivance, by which the required strain is readily produced and maintained unaltered; the retarding friction being quite independent of the condition of the rubbing surfaces. This break was successful, and has been used ever since. The 1858 expedition began operations on the Atlantic, joining the cables contained in the 25th of June by splice in the middle of the fouled the 'Niagara,' and broke. A second Niagara' and Agamemnon.' The cable splice was at once made, and successfully

6

lowered to the bottom. When the Agamemnon had payed out 37 miles, and the 'Niagara' 43 miles, the electrical tests showed that the copper conductor of the cable was severed. In technical language, there was a loss of continuity. The 'Niagara' endeavoured to haul in the cable, which shortly broke for the third time. On the 28th of June another splice was made; but after 111 miles had been paid out, the cable broke for the fourth time, with a strain indicated of 2200 lbs., or nearly one ton. On the 28th of July another splice was made, and this time the cable did not break, but was laid successfully as a mechanical operation, but unsuccessfully in all other senses. As before stated, a want of continuity did occur, but it ceased after a few hours, and was passed over as of insufficient consequence to stop the submersion.

Much surprise has been expressed at the rupture of a cable estimated as strong enough to bear four tons, when the indicator showed only about one ton. It has frequently been suggested that the instrument gave false indications; but there is really little reason for supposing this. The cable was covered by 126 small iron wires, spun into eighteen small strands, the whole cable being only 5-8ths of an inch in diameter. The wire was not galvanized, and rusted very readily. It is most probable that in many places its theoretical strength was very much reduced by this cause.

In 1865 and 1866 the same break and indicator, or dynamometer, as it is sometimes called, were used, but the history of events was widely different. The cable, during submersion, not only escaped fracture, but was not even once strained within a tenth part of its supposed strength. In 1865, the occurrence of a small fault, which would have been far too insignificant to have been detected in 1857 or 1858, caused an attempt to haul back the cable, which was broken by chafing against a projection from the bows of the Great Eastern.' The arrangements in 1865 were by no means perfect. The picking-up gear was defective, and the system of electrial tests faulty, but the paying-out machinery acted admirably, and the cable hardly admitted of improvement. In 1866 the picking-up gear was good, and the electrical arrangements left nothing to be desired.

The special form of cable adopted, in which each iron wire is enveloped in hemp, presents various interesting peculiarities. It is actually stronger than the sum of the strengths of the hemp and steel employed to make it. This almost incredible paradox was discovered during experiments made by

Messrs. Gisborne, Forde, and Siemens for the Government, with reference to a proposed Falmouth and Gibraltar Cable. It seems strange enough that a steel wire can be strengthened by wrapping hemp or manilla round it; but this was soon found to be a fact, and indeed the percentage of elongation undergone by a hempen strand and a steel wire before breaking are by no means so different as most people would imagine. By selecting the best lay of the hemp round the steel, it was repeatedly found that the strength of the two combined exceeded the sum of the strengths of the two separately, and this strange result has been fully confirmed by independent experiments conducted by Mr. Fairbairn and others for the Atlantic and Telegraph Construction Companies. The explanation is simple enough. Neither material is really homogeneous: each has its weak places; it is extremely unlikely that the weak places of both should coincide. When, therefore, the two are combined, we obtain the sum of the average strengths of each material; when they are tested separately, we get the sum of the strengths of the two at their weakest points.

This form of cable was first used in 1860 for a cable between France and Algiers, Messrs. Gisborne and Forde being the engineers, and Messrs. Glass and Elliot the contractors. The cable, after some misadventures, was successfully laid, and behaved well during submersion, but the form fell into some discredit, owing to the discovery that even in 1500 fathoms the hemp was eaten away by a species of teredo after a few months of submersion. This left a mere cage of loose iron or steel wires, unfit to be lifted, or relaid if lifted. Fortunately it appears that these animals, which in the Mediterranean fasten on every inch of exposed hemp, do not exist in the Atlantic. Where they have eaten the hemp, the guttapercha appears as if marked with the smallpox; but no instance has yet occurred where they have actually penetrated the guttapercha to any serious depth.

The form has other defects. Many persons think that the two injuries which the 1865 cable received during submersion were not due to malice, but to short pieces of broken wire, which would penetrate the soft sheathing of hemp with much greater ease than the hard mail of the common iron-covered cable. The arguments used in favour of this view are as follows:-The hemp conceals a break in the wire which it encloses; a broken wire may be bent out when being coiled, and penetrate the neighbouring coil; the injury may not occur, or not be fully completed, until the coils are disturbed by the tramp

ling of the large number of men engaged on the coil when it is being payed out. Pieces of broken wire were found actually sticking out in this manner after attention had been drawn to the possibility by the faults which occurred. Probably, however, the great success of the Atlantic Cables will cause their form to be the type for deep-sea lines for some time to come.

Cables on board ship are now almost invariably stowed in water-tight tanks; from these they pass up to a sheave or quadrant over the centre of the coil, and thence to the break-drum, and over the stern. A turn or twist is put into the rope at every turn which it makes round the tank; that is to say, it is twisted tighter by the mere action of coiling away; but this twist is again taken out when the cable is uncoiled; so that if this operation proceeds with regularity, the cable goes into the sea in the same condition as it left the sheathing-machine; but if the cable is stiff and springy, or if it is drawn from the hold by jerks, or if one or two coils stick together and are drawn up at once, the turn in the cable tends to throw it over into a loop, which may easily be squeezed or drawn into an ugly-looking thing called a kink. With circular coils, and experienced men in the hold, this hardly ever occurs, and it is rendered next to impossible if the eye of the coil is filled up by a smooth cone, to which the rope clings in ascending, and which prevents any coil from being drawn into a loop. This cone, together with certain guiding-rings which prevent the cable from flying out under the action of centrifugal force, form the subject of a patent taken out by Mr. Newall, and first used in 1855 for the Varna-Balaclava Cable. The excellence of the contrivance hardly admits of a doubt; but the action of the Patent Laws receives some curious illustrations from the incidents which this patent has given rise to. The validity of the patent has been greatly contested; substitutes more or less like the thing patented have been devised, but rival manufacturers have seldom consented to use the thing patented, and pay the royalty. Although the holds were arranged with contrivances having the same object as Newall's cone and rings, foul flakes, as they are called, twice came up from the hold, once on each expedition. These foul flakes are simply two or more turns of the cable which come up entangled together, and then get jammed into more or less of a tangle on deck, for round the break drums they cannot go. The cable has to be stopped at once, the ship's engines reversed, and all hands busied in setting the mischief to rights. The following extract from a speech delivered at Glas

gow by Captain Hamilton, who accompanied the expedition as a Director of the Atlantic and Anglo-American Companies, gives a graphic description of the foul flakes which occurred during the laying of the 1866 cable :

'This interruption occurred in consequence of the cable, which was being payed out from the after-tank, bringing up with it a bight from the next lower flake, and also the lead from the inside to the outside of the next layer of the coil, so that five cables were running out from the tank instead of one.

'These were carried aft together till they were stopped by the paying-out machinery; when, in a very short time, they appeared like the tangle of a gigantic fishing-line. The ship was immediately stopped, but the night was pitch dark, rain falling heavily, and a fresh breeze blowing, the cable over the ship's stern being only visible by a slight phosphorescent light where it dipt into the water. Sir James Anderson, however, by great skill, contrived so to handle his ship of 23,000 tons, which was riding at single anchor in 2000 fathoms by a mere thread, that the engineers and sailors had time to reduce this apparent confusion to order, and in about three hours the paying-out was resumed without the perfect testing of the cable having been in the slightest degree inter

fered with.'*

160 or 170 miles of cable were payed out daily during the 1865 Atlantic expedition, and from five and a half to six and a half knots per hour may be considered a good speed in cable-laying. In 1866 the speed was rather slower, the distance was generally about 120 miles per diem, and the cable payed out about 135 miles. The 1865 and 1866 cables are 1896 and 1858 nautical miles long respectively. The total distance from shore to shore is 1670 nautical miles. The 1858 cable was 2022 miles long, and it was payed out as fast as in 1865, but more cable was wasted, and the ship went slower. A footnote gives the principal dimensions and weights of these cables.†

1866.

From the Glasgow Daily Herald, 5th November

↑ First Atlantic.-Length as laid, 2022 knots; copper conductor 7-wire strand, weighing 107 lbs. per knot, diameter 0-083 in.; covered with guttapercha, weighing 260 lbs. per knot, diameter, 0.38 in.; served with tanned hemp, and covered with eighteen strands of seven bright charcoal iron wires 0-028 in. diameter; total diameter of cable 0.62 in.; weight of cable in air per knot 217 cwt.; in water 16.3 cwt.

Second, or 1865 Atlantic.-Length when complete in 1866, 1896 knots; copper conductor 7-wire strand weighing 300 lbs. per knot; diameter 0.114 in.; covered with gutta-percha and Chatterton's 0-464 in.; served with wet tanned hemp covered Compound, weighing 400 lbs. per knot, diameter with ten bright steel wires, each enclosed in five tarred manilla hemp strands, diameter of each wire

« ÀÌÀü°è¼Ó »