strength when made into chain, which is due no doubt to the welding process. Table 1 shows that bar steel is superior to bar iron to the extent of 47·84 ÷ 25.7 = 1.86, or 86 per cent.; but when welded, iron is superior to steel, 21.120.4 = 1.034, or 3.4 per cent.

=

=

When made into stud-linked chain the mean strength of steel by Nos. 38 to 43 of Table 21 is 18.13 tons per square inch, iron being 17.44 tons; practically, therefore, there is little or no difference in the two materials.

=

There are two kinds of chain in common use, the short-linked or crane-chain used for most purposes on land; and the stay or stud-linked cable-chain for naval purposes.

(103.) "Short-linked Crane-chain."-Table 21 shows that the mean strength of crane-chain from inch to 1 inch diameter is 19 tons per square inch, and it appears to be about the same for all the sizes between those extremes. A chain made of 1-inch iron would therefore break with 19 × ·7854 × 2 = 29.84, or say 30 tons, and for short-linked crane-chain we have the rules:

tons W

=

(104.) Mean breaking weight in pounds, w = d2 × 1050. d2 x 47. (105.) Government proof-strain in pounds, p = d2 × 420. tons, P = d x 1875. In which d = the diameter of the iron in 8ths of an inch, &c.:

=

thus for 1-inch chain we have W 64 × 47 30 tons, and P = 64 x 1875 = 12 tons, &c.; cols. 2 and 3 in Table 22 have been calculated by these rules. It will be observed that the ratio of the proof strain to the breaking weight is 1 to 1050 420 = 2.5. With lifts, cranes, &c., where life is jeopardised, the safe working load should not exceed th of the breaking weight, but for many less critical cases it may be 50 per cent. more than that, or ths of the breaking weight: we thus obtain cols. 4 and 5 in Table 22. The weights per fathom (or 6 feet) in col. 6, from inch to 1 inch, were found by weighing given lengths; the rest were calculated from the 1-inch chain.

(106.) "Stud-linked or Cable-chain."-Table 21 shows that the strength of stud-chain is not so great as is commonly

TABLE 22.-Of the STRENGTH and WEIGHT of SHORT-LINKED

supposed: the mean of the twenty-four experiments, Nos. 14 to 37 on iron chains from inch to 2 inches diameter is 17.43 tons per square inch, or 11⁄2 ton less than that of ordinary crane-chain. This is contrary to the current notion on the subject, but is the clear result of experiment: cable-chain has, however, some important advantages, principally in its nonliability to kink or become entangled, which for naval purposes is all-important: moreover it is lighter, as shown by col. 5 of Table 23.

Admitting 17.43 tons per square inch as the mean strength of cable-chain, we have 17.43 × 7854 x 2 = 27.37 tons, the

TABLE 23.-Of the STRENGTH and WEIGHT of STUD-LINKED CABLE

breaking weight of 1-inch chain: taking 27 tons for round numbers, we have for stud-linked chain, the rules:—

(107.) Mean breaking weight in pounds, w d2 × 945.

In which d = the diameter of the iron in 8ths of an inch: thus for 1-inch cable W = 64 × 423 = 27 tons, and P = 64 x .282 = 18 tons, &c.: cols. 2 and 3 in Table 23 have been calculated by these rules.

The ratio of the proof strain to the breaking weight is 1 to 945630 1.5, which is a very severe test, but it is maintained that the object of testing, namely, to discover faulty

=

links, would not be answered without a heavy strain, and that experiments at Portsmouth Dockyard have shown that the strength of a chain is not seriously or even sensibly impaired by repeated strains almost equal to the breaking weight. In ordinary cases, however, it is not desirable to load chain-cables to more than half the Admiralty proof-strain, or 3rd the breaking weight, and from this we obtain col. 4 in Table 23.

(109.) "Annealing Chain."-It is shown in (14) that the effect of annealing wrought iron in the form of bars, plates, or chains is a loss of tensile strength. The effect on chain is very clearly shown by Nos. 19, 20, in Table 21, where the matter was submitted to direct experiment: ten specimens of 7-inch annealed chain gave a mean strength of 16.34 tons per square inch; and ten similar ones not annealed, gave 17.54 tons, showing 16 3417.54 93, or 7 per cent. loss. Comparing the maximum strengths in col. 4, or the minimum ones in col. 5, we obtain similar results: thus with the former we have 20.25 ÷ 21·75 = ·93, or 7 per cent.; and with the latter 1920.5 93, or 7 per cent. also.

=

=

But if in the course of manufacture, the iron is cold-hammered, which not unfrequently occurs in practice, a loss in strength of about 30 per cent. may accrue as shown in (14), and in order to avoid that great loss, it is expedient to submit to the 7 per cent. loss due to annealing.

It has been found by experience that by long-continued use, chains become brittle and break with a comparatively small impulsive strain, such as very often occurs with cranes, &c., by the load slipping or otherwise: thus a g-inch chain has been known to break in one case with 3 tons, and in another with 5.9 tons, although the breaking load by col. 2 of Table 22 should not have been less than 11.72 tons. It should be observed, however, that in both cases the strain exceeded the safe working load given in col. 5, namely, 2.34 tons.

It is very desirable that all chains should be periodically re-annealed and re-tested: the failure of chains is the source of most of the serious accidents in our factories, which might be avoided to a great extent by more frequent annealing and testing.