History of the Atlantic Cable & Undersea Communications
On The Causes Of Failure Of Deep-Sea Cables
Introduction: This article on the failure of the 1865 Atlantic cable is from the Journal of the Society of Telegraph-Engineers and Electricians, No. 51, Vol. XIII, 1884.
James Graves was a cable engineer who supervised the manufacture of the 1865 cable and was from 1866 - 1909 superintendent of the Valentia Cable Station, where the recovered 1865 cable terminated. See Cornelia Connelly's article on the Irish cable stations for more information on Valentia, and this article for a review of Graves's career.
In October, 2002, a memorial to mark the laying of the trans-Atlantic cable from Valentia to Newfoundland was unveiled on Valentia. Made of Valentia slate and designed by sculptor Alan Hall the memorial marks the history of the telegraph industry on the island from 1857 forward. The first Superintendent of the station was James Graves, and his great-great-grandson, Gordon Graves, was the guest of honour at the unveiling.
51 1 8 8 4
The quarterly Journal of the Society (Vol. XII., No. 50) has just reached me, giving in extenso the paper of Messrs. Trott and Hamilton, and the discussion thereon upon the 29th November, 1883, and I am extremely interested therein, although the discussion did not develop the subject brought forward to any great extent, as the speakers appeared to doubt the facts laid before them by the authors.
Many references were made to the 1865 Atlantic cable, which, on behalf of the Atlantic Telegraph Company (as the representative of Mr. C. F. Varley), I tested throughout from the commencement of the core manufacture to the completion of the shipment on board the "Great Eastern," and therefore took a most lively interest in its welfare, not only before it was laid, but afterwards in superintending its working until it died a natural or a suicidal death at about 9 a.m. on the 11th March, 1873, and to my great regret was never resuscitated.
This break, about 580 knots from Valentia, in 1875 fathoms, was followed almost immediately after, on 9th April, 1873, by another about 795 knots from Newfoundland, in 2,400 fathoms, and three years later, on 10th June, 1876, by yet another about 185 knots from Valentia, in 1,200 fathoms. This length remained intact until January, 1882, when it was broken during the repairs to the 1874 cable off the Skelligs. Meanwhile part of the western end of it had been picked up and utilised to form the whole of one and half another cable from Heart's Content across Trinity Bay to Rantem - one of which being termed the "Rantem" cable and the other the "Island Cove" cable, in order to distinguish them. These cables are laid chiefly in depths of 150 to 200 fathoms, and are about 60 miles long, and work admirably.
Such has been the fate of the cable of 1865. The type of this cable was so carefully considered before the specification was completed and adopted, and such diligent care was bestowed throughout its manufacture to secure a perfect cable, that it is worthy of the deepest consideration for the purpose of endeavouring to discover why its laying could not be completed during the summer of 1865, and why it should have failed after it was laid in several places in deep water; that being the type more or less adhered to in all the deep-sea cables since laid, aggregating many thousands of knots, and representing many millions of capital.
As the practical representative of the Atlantic Telegraph Company during the manufacture of this 1865 cable, and also their representative at Valentia during its submersion, as well as their already appointed superintendent at this station, I very naturally sought out the prevailing opinions of the leading men engaged on the "Great Eastern," soon after her return to England, as to the cause of the failure, and from notes made at the time, bearing date 20th September,1865, I select the following extracts:-
No. 1 was afraid the injuries were malicious.
No. 2 thought the paying-out system defective, as during half the time there was no check on the insulation. He was decidedly of opinion that the first and second faults were purely accidental, and upon seeing the first fault remarked, "We shall have plenty of these." He believed the iron wire to be brittle in some places, and to become broken during shipment, unnoticed. He saw one place where six inches of wire was missing from the hemp, and the ends protruding. In coiling down, these brittle scraps get between the flakes, and are squeezed into the cable, but may not penetrate the core until, in passing through the paying-out machinery, they get pressed into it. He suggested softer wire, even to the partial sacrifice of strength.
No. 3 believed faults malicious. He would have had the culprit lynched if he could have found him. He considered the picking-up gear too weak for such deep water. He would have no strangers engaged at the last moment before sailing. Sometimes the engines were not stopped, and at others were not started promptly upon the order being given, and extra strain was thus brought to bear on the cable.
No. 4 was on duty in the tank when the third fault occurred. He heard a scraping noise; saw a broken wire; called out, "A broken wire!" and soon after a fault was reported. He was of opinion the second fault was maliciously done, as one end of the wire was partially pointed by filing, and the other looked as if filed partially through and then snapped off.
No. 5 was of opinion the first two faults were caused maliciously. He examined the piece of wire from the second fault with a lens, and was convinced that one end had been tapered by grinding, and not by filing, and that the other end had been nicked with a file at a length equal to the diameter of the cable, and that this had been forced into the cable and then broken off.
No. 6 believed both faults were malicious, and that they could not possibly be accidental.
No. 7 firmly believed the faults to be malicious.
No. 8 thought the second fault was done maliciously. He recommended that no strange men be allowed on board, and none but well-tried factory men be employed. He saw several strange faces, though he was himself an old employee.
The third fault was never recovered, but there is no doubt that the broken wire seen by No. 4 had something to do with it.
These extracts are given here merely to show the current feeling at the time of the 1865 failure amongst those who were actually engaged upon the work, after both officers and men had had a month or six weeks to ruminate over the bare facts and draw conclusions therefrom.
The most probable causes of these three faults were those given by No. 2, as it is easily imaginable that out of 20,000 miles of wire a few brittle short pieces should exist, fall out of the hempen strands, and bury themselves in the tank; the superincumbent weight squeezing them into the covering of the cable, and then upon passing through the machinery the mischief was finished. By the time the tests showed the fact, and the electricians could verify them before giving the alarm, 10½ knots in one case and 2½ in another had been paid out. Hence, assuming the wire with which the cable was "protected" to be the real "culprit," there is but the one conclusion, that it alone was the cause of the failure to lay this cable successfully in the year 1865; and this led me to suggest, in a letter to the Secretary of the Atlantic Telegraph Company, the total abolition of iron in the covering of the next cable.
Now comes the question, why this cable broke in deep water in several places after it had been completed and worked for several years.
I know the usual assumption that the cable was suspended in catenary curves from peak to peak, as referred to by Mr. Forde (p. 513), owing to inequalities in depth within short distances, and so got broken when sufficiently weakened; but has this been sufficiently proved in relation to this 1865 cable, or is it still mere assumption? because the latter would be poor evidence upon such a vital question, however easy it may be to get out of a difficulty by such an assumption. Should we not rather fall in with the views of Captain Trott and Mr. Hamilton, and attribute these failures in deep water to the "wringing" phenomenon developed by their experience, and brought to bear upon the weak places in the cable?
Professor Fleeming Jenkin, in his Cantor lecture at the Society of Arts, February 5th, 1866, laid great stress upon the protective nature of a completely closed cable sheathing, and showed that it could not stretch because it formed a solid tubing round the core; and when alluding to the Malta-Alexandria cable, he remarked that "no sensible untwisting ever does occur; about 40 or 50 turns are, at most, taken out per mile." And further, referring to the 1865 cable, he says it "is the strongest cable yet made, bearing more than twice as great a length of itself as the old iron cable. The new form stretches more than the old. The hemp may be eaten off, or decay from the wires, weakening the cable, and the hemp affords less mechanical protection against injury, but the stretch is never such as to endanger the core, as has been proved by repeated experiments."
The twisting (or untwisting) of the cable in paying out, although in theory and experiment said to be not "sensible," has, however, been observed by Messrs. Trott and Hamilton in actual practice. To be visible between the stern of the ship and the surface of the water it is fair to assume that two turns would have to be made to attract notice, and that on a length of probably not more than 20 fathoms.
Now, if two turns are taken out in 20 fathoms, and this action is continued in the same ratio to the bottom, at a distance of 2,000 fathoms, it would amount to 200 turns: how much more would it be when, as Mr. Forde remarked (p. 513), in such a depth there are frequently "20 or 30 miles of cable probably suspended in the water from the stern of the ship to where the cable touches the bottom"?
Or, to take the more modest estimate of Professor F. Jenkin, say, 30 turns per mile (50 being the possible maximum), and applying his further estimate, that in 2,000 fathoms, the cable being laid from the ship at an angle of 9º 30' (average) and a strain of 12 cwt. (average), there are from 12 to 13 miles off the bottom, 30 turns per mile for 12 miles would yield 360 turns.
Our late President said that he "could not say how many twists a length of core would stand before breaking, because he had never had patience to go far enough; but he had twisted it 50 times without obtaining a break." This was, as a matter of course, tried upon new core. What would be the probable effect of 360 or more turns on old core?
As regards the protecting power of the iron, it is well known that it does not form a solid tube round the core, so as to resist the crushing power of a heavy strain in the Atlantic type. The diameter of the steel wires (homogeneous) is 0.095 inch, and these, being 10 in number, would form a circle whose circumference would measure (through the centre of the wires) 0.95 inch; and as the diameter of the core is 0.464, its circumference would be 1.45 inch, leaving a difference of half an inch in the event of all the hemp and jute being destroyed and removed. The ten wires at their full gauge, when new, would not surround the core in a complete helical tube, and any twist or strain would tend to cause them to cut into the core and damage it.
In paying out a cable, the twists are spread out and kept stretched as long as a steady strain is kept upon the cable, and are distributed along its length; but if the ship stops, and this strain is relieved and a lot of slack runs out, as it will do, a reaction sets in, and the twists which have been taken out of the length remaining off the bottom immediately endeavour to recover themselves, and expend their force upon the slack at the bottom, and form hoops, ready to be drawn into kinks the moment a strain is put upon them in the attempt to pick up, and then a fracture results. Every one knows how much easier it is to break a piece of wire by putting a kink in it, than by fairly straining it. Hence the cable breaks "at the bottom," and "not near the ship," "when" (p. 528) "the ship is still and the cable hanging perpendicular."
It appears to me that this can only be avoided by keeping a permanent strain upon the cable, as in paying out, but in practice this is too difficult a thing to ever hope to do on board a ship.
Again, as regards the twisting effect in picking up a cable in deep water, these twists, or rather untwists, are driven from the bow-sheave towards the bottom. If the cable could be hauled in hand over hand by manual labour, without any pressure upon it, the cable might stand a chance of coming up with its lay intact, but in practice this is impossible. The cable lies in the sheave, and the depth of water and the specific gravity of the cable regulate the pressure between the cable and the groove of the sheave. This pressure is sometimes very great. Captain Trott remarked (p. 527), "I have seen the cable come up quite flat over the sheave, the lay much elongated, and the core spewing out." This pressure makes the section of the cable a flattened oval, instead of a circle; the lay cannot pass this great pressure into the ship as it should do if no turns were to be taken out; consequently the lay is lengthened considerably by the pressure and the strain of hauling it in by steam power over the sheave, and thus the turns taken out of the lay on the sheave are continually forced away from the ship towards the bottom, or to the weakest "link in the chain" off the bottom, and there expend their force (possibly amounting to several hundred turns in all) upon the weak spot, to the wrenching asunder of the fabric which, as Mr. Hamilton said (p. 531), "occurs invariably when cables are picked up in deep water." Of course he refers to cables crippled by age, as new cables containing their full element of strength have been known to be picked up without such accidents.
The most feasible explanation of the failures in deep water of submarine cables is, that from some geological cause or other at the bottom of the ocean the iron wires become weakened by oxidation, the hemp as a consequence becomes destroyed, and that the accumulated twists on both sides of a weak place concentrate their force upon it and wring the cable asunder, and that these failures are not necessarily due to abrasion on elevated ridges.
If therefore a cable can be made, such as has been suggested by Captain Trott and Mr. Hamilton, without twists, - one which "cannot twist" (p. 526), - there is some hope that such accidents and failures may be avoided in the future.
Valentia, February 12, 1884.
Return to the Atlantic Cables index page
Last revised: 12 June, 2016