bars, and it is to be greatly desired that of plate smiths or when built into struc similar observations should be made on tures, they behave very differently to all kinds of modern metallic structural iron with the same apparent mechanical materials. The mild bars to which Dr. properties. The toughness of good Siemens' table refers shows a mean elas- iron plates enables them gradually to tic limit of 17.37 tons when annealed, dissipate any internal differential molecuthough the same bars showed the lower lar strains of tension and compression mean elastic limit of 16.62 before being that may be resident in them as they annealed, and curiously the elastic ex- leave rolling mills, by differential comtension is also greater in the annealed pression and tension. Steel plates bars, the mean extension in bars 5 feet which are comparatively hard, but which, 0 inch, and 4 feet 11 inches, being 0.086 when torn asunder in test-strips, indiinches, the value of the Te of the bars cate an ultimate extension of as much being by the before-mentioned formulæ, as 15 per cent., might be expected to do respectively 57.04 and 53.348. This de- the same. Such, however, is not the notes a high structural value, although case, for the plate has often behaved, this indication must be taken with the when built up into a structure, as though value of Tr which cannot be gained its ultimate extension was not more than from the table referred to. Such mate- one or two per cent., and, like glass, posterial, however, is not to be had or can- sessed of a high elastic limit, but no not be safely used in plates, and even in toughness. Why this should be is not bars has been very little used in bridge- known, but it may be suggested that work. Steelmakers have yet to satisfy such being the behavior, the following engineers that it can be safely used. may afford some clue to the fact that thick plates, at least of such material, have fractured in various directions, after the structure of which they have formed an integral part has been completed. Most plates before being built into a structure are annealed, but the following remarks apply equally whether annealed or not.

The most of the recorded experimental results of tests of mild steels do not give the extension of the elastic limit, so comparisons with other materials cannot at present be made. From what has been said, however, it will have been seen, that it is very necessary that these figures should be obtained, and what has often been observed may be here repeated, namely, that in order to the production of a satisfactory series of tables of the mechanical properties of different structural metals, experiments should be conducted on a uniform basis, and with uniform lengths of tests pieces, or if not of uniform length, they should be of a minimum length of either ten inches or one foot.

Such experiments would be perhaps costly, but there is at least one wealthy engineering society, by a committee of which such experiments might be usefully carried out on the structural mateterials of to-day.

In conclusion, a few words may be said upon the behavior of the harder kinds of steel plates in which they differ from those of iron. It has been observed by many that some of the steel plates with an elastic strength as high as 18 tons per square inch combined with a ductile extension as high as 15 per cent., as obtained by the tests of strips of such plates, that in the hands

Plates when taken from the rolls or from the annealing oven are generally laid on a flat surface to cool, but whether laid down or stood on edge, cooling takes place somewhat more rapidly towards the corners and edges than at the middle. At first, the whole plate is of the same temperature, which may be that of redness. The exterior parts first assume the rigidity of cold iron, and contraction takes place on the interior parts which remain at a higher temperature, and therefore the contraction has taken place under a tensile strain. Thus, if a plate of 1 inch in thickness is considered in illustration, it will be seen that a corner of, say 6 inches on either edge, has an area on the two sides of 36 square inches, but it has also the additional effective cooling area of the edges, which adds 12 square inches, making for a surface area of 36 square inches a total of 48 square inches of effective cooling area by radiation and evection. If, on the other hand, an area on the two sides at the center of the plate, of 36

square inches of such surface, be taken these internal strains are aggravated, into consideration, it will be seen that and they may possibly be of such magnithe edge surface can only be considered tude that extraneous strains, that would as cooling by conduction. Thus, the not materially affect a tough iron plate, effective cooling area of the outer parts may be sufficient in a hard steel to cause of the plates is much more efficient than rupture. Further, when a plate of such the central parts. These outer parts a character is being riveted up, every having, then, become rigid and con- rivet is compressed under a very high tracted under tension, exert a corre- strain to make it fill the holes, and thus, sponding compressive strain upon the acting as a viscous fluid, adds to the interior parts still at higher tempera- strains already tending to destroy the ture and thus more or less amenable to plate. compression. This tensile strain upon These remarks are only made as sugthe outer parts or borders of the plate gestions, and should perhaps have been is gradually eliminated as the interior put in the form of a question, as they parts cool, and is finally changed into are somewhat aside the object of this one of compression, as the inner parts note, which is to invite discussion on contract in cooling under a molecular what seems to be the tendency in the tensile strain, due to the incapacity of production of very mild steel plates, the rigid border to follow the inner parts namely, that in the endeavor to remove in their contraction. In the cold plate the difficulties which have attended the put into a structure, there is thus initial use of steel plates of high elastic limit, molecular strain differentiating from com- we seem to be in danger of losing the pression at the edges to tension towards facilties for producing lighter structures the center. If the plate is cooled under which the application of steel seemed circumstances inducing unequal cooling at one time to afford.

THE STRENGTH OF BOILER FLUES.

From "The Engineer."

VERY high pressures are now carried it. The pressure was to be found by the at sea on the outside of tubes of com

2st d

where s is the

which the iron is to be subjected, t the
strain in pounds per square inch to
thickness of the plate, d the diameter of
the boiler, and p the working pressure.
half that which would suffice for a shell.
Then the proper thickness for flues was
This practice is, we are happy to say, no
while moderate pressures not exceeding
longer followed. It answered fairly well
about 45 lbs. on the square inch were
used, but it was totally unfit to deal
with such pressures as are now carried
at sea and on land.
following formula for the strength of
Wilson gives the

paratively large size. These tubes are well-known rule p= the cylindrical furnaces of marine boilers, and reach in some cases a diameter of 4 ft. It is of the utmost importance that very definite rules should be laid down for the guidance of those who design such furnaces, in order that no mistakes may be made. We say "definite rules," because the whole subject has been fully investigated. There is appar; ently nothing more to learn about it, and there should, therefore, be no trouble in constructing a simple formula which would enable an engineer to tell, with very little calculation, what is the proper thickness for a furnace tube of any given length and diameter, intended to sustain a stated pressure. Something of the kind will be found in almost all treatises lapsing pressure in pounds per square on steam boilers. The practice thirty inch, t the thickness of tube in thirtyyears ago was to treat the flues as though seconds of an inch, 7 the length in feet, the pressure were to be exerted inside of and d the diameter in quarter feet;

flues: p=

262.4 × t
lxd

where p is the col

other rules may be found in other haves very vexatiously and insists on a treatises. Fairbairn has shown that the much lower pressure being carried than strength of a flue to resist collapsing is necessary. Before any opinion on pressure varies directly as the 2.19 this point can be properly pronounced power of the thickness, and inversely as it is necessary to call in some other the diameter and length. Thus: P=33.6×(100) 2.19÷L×d

and p=5.6×(100) 2.19÷Ld. These are very ungainly formulæ, and useless without logarithms. It may not be superfluous to say here, however, for the guidance of those who would like to use Fairbairn's rules as a check on their own practice, that in using the formulæ the thickness of the plate is to be multiplied by 100. The log. of the result is to be found and multiplied by 2.19. This gives a log. the natural number of which is the 2.19 power of 100 t.

authority. Referring to the tables given in Wilson's treatise on steam boilers, we find that the collapsing pressure of a flue 7 ft. long, inch thick, and 40 inches in diameter is nearly 400 lbs. on the square inch. Lloyd's factor of safety is consequently about 9 to 1, while the Board of Trade factor is nearly 15 to 1. If one margin be enough then the other must be too great.. In dealing with this part of the question we can only arrive at anything like a satisfactory conclusion by resorting to the result of experiment. Mr. Wilson's figures refer to very perfect tubes, such as may or It is not to be supposed that the may not be met with in practice. Mr. Board of Trade, which is so precise in D. K. Clark has investigated many cases its instructions and rules for marine of collapsed tubes, and he has prepared engine builders, would allow this subject the following formula:

to pass without consideration. Accordingly the Board has proposed a rule,

60,000 t

which runs thus: p = (l+1) x d'

60,000 is a constant for furnace tubes with longitudinal seams, lapped joints, and punched holes, single riveted; is the length of the furnace in feet; d is the diameter in inches; and p the working pressure. To illustrate the application of this rule, let us suppose that a furnace is 40 inches in diameter and 7 ft. long, and that the plates are .375 inches 60,000X.140625

Here Applying this rule to the case stated, we have 105.4 lbs. as the collapsing pressure. In this case the Board of Trade factor of safety is nearly 4 to 1, and Lloyd's factor is a little over 2 to 1. It must be remembered, however, that Mr. Clark's rule applies to flues of considerable length without any strengthening rings; to these he has attached no precise value. But it may be taken for granted that a marine boiler furnace tube well secured at each end and not more than a few feet long, is very much better able to stand up against a collapsing strain than a tube 25 ft. or 30 ft. long. We may, we believe, take a mean between Mr. Wilson s figures and Mr. 89,600 Clark's, and assume that the collapsing have a rule also, which is, p= lxd pressure of our tube would be about 200 Applying this rule to the furnace whose lbs. on the square inch. Under these dimensions we have just stated, we have circumstances Lloyd's formula gives a factor of safety of 4.4 to 1, while the 89,600 ×.140625 Board of Trade rule gives 7.7. 7×40

thick. Then

=26.3 lbs.

(7+1)×40 as the working pressure. But the Board of Trade rule is not the only one with which engineers have to deal. "Lloyd's"

=45 lbs. We thus find

Now, it is evident that the disparity that one great authority on marine en- between the rules of the Board of Trade gineering allows nearly twice as great a and of Lloyd's ought not to exist. It pressure to be carried as the other. If places engineers and shipowners alike in the Board of Trade be right, then a very unpleasant position, and will Lloyd's must be wrong, and dangerously some day cause a good deal of trouble. wrong. If, on the other hand, Lloyd's A case may be cited which recently ocare right, then the Board of Trade be- curred. Two wing furnaces on board a

North-county steamer collapsed with 20 ing spectacle presented for consideralbs. of steam pressure and plenty of tion.

water. The collapsed flues are round The engineers of a great public detopped, with flat stayed sides, and by partment asserting that a boiler is the Board of Trade rules, the working strong enough, while those of the Govpressure was 22 lbs. and by Lloyd's 42 ernment assert that it is too weak, will ĺbs. It is probable that when the case not be a satisfactory display in any comes to be investigated it will be found sense of the word. The question at that the metal was either over-heated by issue is one really of very great importthe presence of deposit, or that the flues ance to engineers and shipowners. Is it were worn. Be this as it may, it is not too much to ask in the interests of cominconceivable that the engineers of the mon sense that the two bodies should Board of Trade and those of Lloyd's put themselves in communication and may some day come into collision in a agree on so apparently simple a matter court of law over such questions, and we as the preparation of a rule for the shall then have anything but an edify-thickness of furnace tubes?

INFLUENCE OF TEMPERATURE IN TUNNELING THROUGH HIGH MOUNTAINS.

By DR. F. M. STAPFF.

Translated from Revue universelle des Mines, for Abstracts of Institution of Civil Engineers.

As the difficulties encountered from keeping perfectly still and in quite pure elevated temperature in tunneling air. In the author's experience on minthrough high mountains require the ing and railway works in Mexico and the adoption of appliances and resources south of the United States the temperavery different from those hitherto em- ture was as high as 40° C. or 104° F. ployed, the author, as Engineering In tropical seas the temperature in the Geologist of the St. Gothard Tunnel, stoke-holes of steamers occasionally Airolo, has during the past six years been collecting voluminous materials for the purpose of endeavoring to solve the two following questions:

1st. What is the highest temperature at which men can work underground?

2nd. At what depth below surface is this temperature likely to be reached in tunneling?

reaches even 69° C. or 156° F., and is aggravated by the dust raised in the act of stoking. Unfortunately there is not sufficient information at present available as to the practicable limits of temperature in underground workings. In upcast shafts used for winding, and for raising and lowering the men, a temperature of from 27° to 32° C. or 80° to 90° I.-Temperature at which underground F. is allowed for the air current in work becomes impossible. The limit of English collieries; and in Belgian from heat at which men can work depends 22° to 34° C. or 72° to 94° F. At upon the length of their exposure to it, Fahlun copper mine in Sweden, in rethe amount of exertion they put forth, opening some old workings that had their acclimatization, the nature of the been stopped by a fire twenty years atmosphere, and, most of all, its degree previously, the clearing up was done at of moisture. Omitting instances of a temperature as high as 52° C or 125° momentary exposure to exceptional heat, F.; but there the natural ventilation was it is certain that men cannot accustom excellent. In another mine, stoping was themselves to stand for any length of abandoned at a temperature of 33° C. time more than 60° to 75° Centigrade or or 91° F., where the dust from the 140° to 165° Fahrenheit, even when decomposed pyrites was more intoler

perature be still further lowered, but also increase of moisture would be prevented. The men would have to work short shifts, and the heat of the body should not reach 40° C. or 104° F., the ordinary bodily temperature being from 363° to 38° C. or 98 to 100° F.

able than the mere heat, and dyed the cious if covered with a layer of common miners as black as ink. The highest salt, whereby not only would the temtemperature observed in the Mont Cenis tunnel was 30° C. or 86° F., at about 4 miles in from the south end. In the St. Gothard tunnel full work was carried on at about 31° C. or 87° F. on the south (Airolo) side, in air surcharged with moisture; and on the north (Göschenen) side at 29° C. or 84° F., in an atmosphere not quite saturated. The author gives detailed particulars of the air-supply to each end of the tunnel, and of the number of men, animals, and lamps consuming it; and on the assumption that the dryness of the air is of equal importance with its purity, he deduces the ratio of 4 to 3 as about representing the superior ventilation of the northern section over the southern at the time of his observations, when each end had been driven about 4 miles in.

The Author goes minutely into the effects observed to be produced upon the health of the men employed in the St. Gothard tunnel. He also reports at length a number of observations he made, as to how hot the men got, and how much work they could do, according to the state of the air in the tunnel. The diminished effort instinctively exerted by them, in proportion as they get hotter from the heat and badness of the air they breathe in such situations, is shown to be in accordance with physiological princiAccording to Professor Dubois-Rey- ples. Taking as the unit of bodily exermond of Berlin, men can stand 50° C. or tion the amount of effort put forth in 122° F. when the air is as dry as possi- merely walking along the tunnel at an ble, as in the case of blast-furnace work- easy pace-say about 4,550 foot-lbs. per ers; but in an atmosphere saturated minute-the Author considers this is with moisture even 40° C. or 104° F. somewhat exceeded by the ordinary exwould almost certainly prove fatal. Air ertion of the workmen employed in the feels very dry when only one quarter roomier portions of the tunnel, behind saturated, but becomes stifling before the advanced heading. Those engaged, complete saturation is reached; hence a however, in loading the broken rock into slight diminution of moisture may be of the wagons near the forebreast, in the the utmost value, and the principal confined space of the advanced heading means of rendering high temperatures itself, must in his opinion, judging from endurable would consist in drying the the shorter time they work and the higher air. A wagon-load of quicklime, not- rate of their wages, be working about withstanding the heat evolved in quenching, he considers would even give a fresher feeling than one of ice, on account of the latter rendering the air still more moist. If at the working place there were both a wagon of lime and one of ice, and also a supply of fresh air direct from the tunnel mouth, he is of opinion that work could be carried on even where the temperature of the rock was as high as 50° C. or 122°F. The ice would indeed be far more effica

twice as hard; and occasional intermittent efforts may be even four times as great. They have to fill the broken stone into baskets upon long narrow lorries running on temporary rails of 1 foot guage, laid alongside the main rails of 34 feet guage; the lorries have to be run back by hand through a distance averaging about 80 yards to the main trucks, into which the baskets have then to be emptied. A score of men formed the gang occupied in this heaviest work. The relative cost of the clearing per cubic *In the last working of the United copper mines in yard of rock in the two headings is found Gwennap, Cornwall, a hot spring of water at 115° F. was met with in the bottom level about 1,500 feet below sur- to be represented by the very same ratio face, and the rock was so hot that even with their thick clothing the miners could not lean against it. (See Pro- as the absolute moisture of the air in ceedings Institution of Mechanical Engineers, 1873, p. 110.) In the Comstock silver mines, Nevada, the tempera- them, namely, 1 to 1.18 in the Göschenen ture averages 130° F. at about 2,000 feet below surface, and and Airolo ends respectively.

sometimes reaches 139° F., while in particular places even 157° F. has been observed. (See Transactions American Institute of Mining Engineers, vol. vii., pp. 45-76; and Minutes of Proceedings Institution Civil Engineers, vol. lvii., p. 393).

The highest limit theoretically possible for the air temperature in tunnel workings would be such as would induce fever