_A._--The iron of boilers, like the iron of machines or structures, is capable of withstanding a tensile strain of from 50,000 to 60,000 lbs. upon every square inch of section; but it will only bear a third of this strain without permanent derangement of structure, and it does not appear expedient in any boiler to let the strain exceed 4,000 lbs. upon the square inch of sectional area of metal, especially if it is liable to be weakened by corrosion.

299._Q._--Have any experiments been made to determine the strength of boilers?

_A._--The question of the strength of boilers was investigated very elaborately a few years ago by a committee of the Franklin Inst.i.tute, in America, and it was found that the tenacity of boiler plate increased with the temperature up to 550, at which point the tenacity began to diminish.

At 32, the cohesive force of a square inch of section was 56,000 lbs.; at 570, it was 66,500 lbs.; at 720, 55,000 lbs.; at 1,050, 32,000 lbs.; at 1,240, 22,000 lbs.; and at 1,317, 9,000 lbs. Copper follows a different law, and appears to be diminished in strength by every addition to the temperature. At 32 the cohesion of copper was found to be 32,800 lbs. per square inch of section, which exceeds the cohesive force at any higher temperature, and the square of the diminution of strength seems to keep pace with the cube of the increased temperature. Strips of iron cut in the direction of the fibre were found to be about 6 per cent. stronger than when cut across the grain. Repeated piling and welding was found to increase the tenacity of the iron, but the result of welding together different kinds of iron was not found to be favorable. The accidental overheating of a boiler was found to reduce the ultimate or maximum strength of the plates from 65,000 to 45,000 lbs. per square inch of section, and riveting the plates was found to occasion a diminution in their strength to the extent of one third. These results, however, are not precisely the same as those obtained by Mr. Fairbairn.

300. _Q._--What were the results obtained by him?

_A._--He found that boiler plate bore a tensile strain of 23 tons per square inch before rupture, which was reduced to 16 tons per square inch when joined together by a double row of rivets, and 13 tons, or about 30,000, when joined together by a single row of rivets. A circular boiler, therefore, with the ends of its plates double riveted, will bear at the utmost about 36,000 lbs. per square inch of section, or about 12,000 lbs.

per square inch of section without permanent derangement of structure.

301. _Q._--What pressure do cylindrical boilers sustain in practice?

_A._--In some locomotive boilers, which are worked with a pressure of 80 lbs. upon the square inch, the thickness of the plates is only 5/16ths of an inch, while the barrel of the boiler is 39 inches in diameter. It will require a length of 3.2 inches of the boiler when the plates are 5/16ths thick to make up a sectional area of one square inch, and the separating force will be 39 times 3.2 multiplied by 80, which makes the separating force 9,984 lbs., sustained by two square inches of sectional area--one on each side; or the strain is 4,992 lbs. per square inch of sectional area, which is quite as great strain as is advisable. The accession of strength derived from the boiler ends is not here taken into account, but neither is the weakening effect counted that is caused by the rivet holes. Some locomotives of 4 feet diameter of barrel and of 3/8ths iron have been worked to as high a pressure as 200 lbs. on the inch; but such feats of daring are neither to be imitated nor commended.

302._Q._--Can you give a rule for the proper thickness of cylindrical boilers?

_A._--The thickness proper for cylindrical boilers of wrought iron, exposed to an internal pressure, may be found by the following rule:--multiply 2.54 times the internal diameter of the cylinder in inches by the greatest pressure within the cylinder per circular inch, and divide by 17,800; the result is the thickness in inches. If we apply this rule to the example of the locomotive boiler just given, we have 39 x 2.54 x 62.832 (the pressure per circular inch corresponding to 80 lbs. per square inch) = 6224.1379, and this, divided by 17,800, gives 0.349 as the thickness in inches, instead of 0.3125, or 5/16ths, the actual thickness. If we take the pressure per square inch instead of per circular inch, we obtain the following rule, which is somewhat simpler:--multiply the internal diameter of the cylinder in inches by the pressure in pounds per square inch, and divide the product by 8,900; the result is the thickness in inches. Both these rules give the strain about one fourth of the elastic force, or 4,450 lbs. per square inch of sectional area of the iron; but 3,000 lbs. is enough when the flame impinges directly on the iron, as in some of the ordinary cylindrical boilers, and the rule may be adapted for that strain by taking 6,000 as a divisor instead of 8,900.

303. _Q._--In marine and wagon boilers, which are not of a cylindrical form, how do you procure the requisite strength?

_A._--Where the sides of the boiler are flat, instead of being cylindrical, a sufficient number of stays must be introduced to withstand the pressure; and it is expedient not to let the strain upon these stays be more than 3,000 lbs. per square inch of section, as the strength of internal stays in boilers is generally soon diminished by corrosion. Indeed, a strain at all approaching that upon locomotive boilers would be very unsafe in the case of marine boilers, on account of the corrosion, both internal and external, to which marine boilers are subject. The stays should be small and numerous rather than large and few in number, as, when large stays are employed, it is difficult to keep them tight at the ends, and oxidation of the sh.e.l.l follows from leakage at the ends of the stays. All boilers should be proved, when new, to twice or three times the pressure they are intended to bear, and they should be proved occasionally by the hand pump when in use, to detect any weakness which corrosion may have occasioned.

304._ Q._--Will you describe the disposition of the stays in a marine boiler?

_A._--If the pressure of steam be 20 lbs. on the square inch, which is a very common pressure in tubular boilers, there will be a pressure of 2,880 lbs. on every square foot of flat surface; so that if the strain upon the stays is not to exceed 3,000 lbs. on the square inch of section, there must be nearly a square inch of sectional area of stay for every square foot of flat surface on the top and bottom, sides, and ends of the boiler. This very much exceeds the proportion usually adopted; and in scarcely any instance are boilers stayed sufficiently to be safe when the sh.e.l.l is composed of flat surfaces. The furnaces should be stayed together with bolts of the best sc.r.a.p iron, 1-1/4 inch in diameter, tapped through both plates of the water s.p.a.ce with thin nuts in each furnace; and it is expedient to make the row of stays, running horizontally near the level of the bars, sufficiently low to come beneath the top of the bars, so as to be shielded from the action of the fire, with which view they should follow the inclination of the bars. The row of stays between the level of the bars and the top of the furnace should be as near the top of the furnace as will consist with the functions they have to perform, so as to be removed as far as possible from the action of the heat; and to support the furnace top, cross bars may either be adopted, to which the top is secured with bolts, as in the case of locomotives, or stays tapped into the furnace top, with a thin nut beneath, may be carried to the top of the boiler; but very little dependence can be put in such stays as stays for keeping down the top of the boiler; and the top of the boiler must, therefore, be stayed nearly as much as if the stays connecting it with the furnace crowns did not exist.

The large rivets pa.s.sing through thimbles, sometimes used as stays for water s.p.a.ces or boiler sh.e.l.ls, are objectionable; as, from the great amount of hammering such rivets have to receive to form the heads, the iron becomes crystalline, so that the heads are liable to come off, and, indeed, sometimes fly off in the act of being formed. If such a fracture occurs between the boilers after they are seated in their place, or in any position not accessible from the outside, it will in general be necessary to empty the faulty boiler, and repair the defect from the inside.

305. _Q._--What should be the pitch or numerical distribution of the stays?

_A._--The stays, where the sides of the boiler are flat, and the pressure of the steam is from 20 to 30 lbs., should be pitched about a foot or 18 inches asunder; and in the wake of the tubes, where stays cannot be carried across to connect the boiler sides, angle iron ribs, like the ribs of a ship, should be riveted to the interior of the boiler, and stays of greater strength than the rest should pa.s.s across, above, and below the tubes, to which the angle irons would communicate the strain. The whole of the long stays within a boiler should be firmly riveted to the sh.e.l.l, as if built with and forming a part of it; as, by the common method of fixing them in by means of cutters, the decay or accidental detachment of a pin or cutter may endanger the safety of the boiler. Wherever a large perforation in the sh.e.l.l of any circular boiler occurs, a sufficient number of stays should be put across it to maintain the original strength; and where stays are intercepted by the root of the funnel, short stays in continuation of them should be placed inside.

BOILER EXPLOSIONS.

306. _Q._--What is the chief cause of boiler explosions?

_A._--The chief cause of boiler explosions is, undoubtedly, too great a pressure of steam, or an insufficient strength of boiler; but many explosions have also arisen from the flues having been suffered to become red hot. If the safety valve of a boiler be accidentally jammed, or if the plates or stays be much worn by corrosion, while a high pressure of steam is nevertheless maintained, the boiler necessarily bursts; and if, from an insufficiency of water in the boiler, or from any other cause, the flues become highly heated, they may be forced down by the pressure of the steam, and a partial explosion may be the result. The worst explosion is where the sh.e.l.l of the boiler bursts; but the collapse of a furnace or flue is also very disastrous generally to the persons in the engine room; and sometimes the sh.e.l.l bursts and the flues collapse at the same time; for if the flues get red hot, and water be thrown upon them either by the feed pump or otherwise, the generation of steam may be too rapid for the safety valve to permit its escape with sufficient facility, and the sh.e.l.l of the boiler may, in consequence, be rent asunder. Sometimes the iron of the flues becomes highly heated in consequence of the improper configuration of the parts, which, by retaining the steam in contact with the metal, prevents the access of the water: the bottoms of large flues, upon which the flame beats down, are very liable to injury from this cause; and the iron of flues thus acted upon may be so softened that the flues will collapse upward with the pressure of the steam. The flues of boilers may also become red hot in some parts from the attachment of scale, which, from its imperfect conducting power, will cause the iron to be unduly heated; and if the scale be accidentally detached, a partial explosion may occur in consequence.

307. _Q._--Does the contact of water with heated metal occasion an instantaneous generation of steam?

_A._--It is found that a sudden disengagement of steam does not immediately follow the contact of water with the hot metal, for water thrown upon red hot iron is not immediately converted into steam, but a.s.sumes the spheroidal form and rolls about in globules over the surface. These globules, however high the temperature of the metal may be on which they are placed, never rise above the temperature of 205, and give off but very little steam; but if the temperature of the metal be lowered, the water ceases to retain the spheroidal form, and comes into intimate contact with the metal, whereby a rapid disengagement of steam takes place. If water be poured into a very hot copper flask, the flask may be corked up, as there will be scarce any steam produced so long as the high temperature is maintained; but so soon as the temperature is suffered to fall below 350 or 400, the spheroidal condition being no longer maintainable, steam is generated with rapidity, and the cork will be projected from the mouth of the flask with great force.

308. _Q._--What precautions can be taken to prevent boiler explosions?

_A._--One useful precaution against the explosion of boilers from too great an internal pressure, consists in the application of a steam gauge to each boiler, which will make the existence of any undue pressure in any of the boilers immediately visible; and every boiler should have a safety valve of its own, the pa.s.sage leading to which should have no connection with the pa.s.sage leading to any of the stop valves used to cut off the connection between the boilers; so that the action of the safety valve may be made independent of the action of the stop valve. In some cases stop valves have jammed, or have been carried from their seats into the mouth of the pipe communicating between them, and the action of the safety valves should be rendered independent of all such accidents. Safety valves, themselves, sometimes stick fast from corrosion, from the spindles becoming bent, from a distortion of the boiler top with a high pressure, in consequence of which the spindles become jammed in the guides, and from various other causes which it would be tedious to enumerate; but the inaction of the safety valves is at once indicated by the steam gauge, and when discovered, the blow through valves of the engine and blow off c.o.c.ks of the boiler should at once be opened, and the fires raked out. A cone in the ball of the waste steam pipe to send back the water carried upward by the steam, should never be inserted; as in some cases this cone has become loose, and closed up the mouth of the waste steam pipe, whereby the safety valves being rendered inoperative, the boiler was in danger of bursting.

309. _Q._--May not danger arise from excessive priming?

_A._--If the water be carried out of the boiler so rapidly by priming that the level of the water cannot be maintained, and the flues or furnaces are in danger of becoming red hot, the best plan is to open every furnace door and throw in a few buckets full of water upon the fire, taking care to stand sufficiently to the one side to avoid being scalded by the rush of steam from the furnace. There is no time to begin drawing the fires in such an emergency, and by this treatment the fires, though not altogether extinguished, will be rendered incapable of doing harm. If the flues be already red hot, on no account must cold water be suffered to enter the boiler, but the heat should be maintained in the furnaces, and the blow off c.o.c.ks be opened, or the mud hole doors loosened, so as to let all the water escape; but at the same time the pressure must be kept quite low in the boiler, so that there will be no danger of the hot flues collapsing with the pressure of the steam.

310. _Q._--Are plugs of fusible metal useful in preventing explosions?

_A._--Plugs of fusible metal were at one time in much repute as a precaution against explosion, the metal being so compounded that it melted with the heat of high pressure steam; but the device, though ingenious, has not been found of any utility in practice. The basis of fusible metal is mercury, and it is found that the compound is not h.o.m.ogeneous, and that the mercury is forced by the pressure of the steam out of the interstices of the metal combined with it, leaving a porous metal which is not easily fusible, and which is, therefore, unable to perform its intended function.

In locomotives, however, and also in some other boilers, a lead rivet is inserted with advantage in the crown of the fire box, which is melted out if the water becomes too low, and thus gives notice of the danger.

311. _Q._--May not explosion occur in marine boilers from the acc.u.mulation of salt on the flues?

_A._--Yes, in marine boilers this is a constant source of danger, which is only to be met by attention on the part of the engineer. If the water in the boiler be suffered to become too salt, an incrustation of salt will take place on the furnaces, which may cause them to become red hot, and they may then be collapsed even by their own weight aided by a moderate pressure of steam. The expedients which should be adopted for preventing such an acc.u.mulation of salt from taking place within the boiler as will be injurious to it, properly fall under the head of the management of steam boilers, and will be explained in a subsequent chapter.

CHAPTER VI.

PROPORTIONS OF ENGINES.

STEAM Pa.s.sAGES.

312. _Q._--What size of orifice is commonly allowed for the escape of the steam through the safety valve in low pressure engines?

_A._--About 0.8 of a circular inch per horse power, or a circular inch per 1-1/4 horse power. The following rule, however, will give the dimensions suitable for all kinds of engines, whether high or low pressure:--multiply the square of the diameter of the cylinder in inches by the speed of the piston in feet per minute, and divide the product by 375 times the pressure on the boiler per square inch; the quotient is the proper area of the safety valve in square inches. This rule of course supposes that the evaporating surface has been properly proportioned to the engine power.

313. _Q._--Is this rule applicable to locomotives?

_A._--It is applicable to high pressure engines of every kind. The dimensions of safety valves, however, in practice are very variable, being in some cases greater, and in some cases less, than what the rule gives, the consideration being apparently as often what proportions will best prevent the valve from sticking in its seat, as what proportions will enable the steam to escape freely. In Bury's locomotives, the safety valve was generally 2-1/2 inches diameter for all sizes of boiler, and the valve was kept down by a lever formed in the proportion of 5 to 1, fitted at one end with a Salter's balance. As the area of the valve was 5 square inches, the number of pounds shown on the spring balance denoted the number of pounds pressure on each square inch of the boiler.

314. _Q._--Is there only one safety valve in a locomotive boiler?

_A._--There are always two.

315. _Q._--And are they always pressed down by a spring balance, and never by weights?

_A._--They are never pressed down by weights; in fact, weights would not answer on a locomotive at all, as they would jump up and down with the jerks or jolts of the train, and cause much of the steam to escape. In land and marine boilers, however, the safety valve is always kept down by weights; but in steam vessels a good deal of steam is lost in stormy weather by the opening of the valve, owing to the inertia of the weights when the ship sinks suddenly in the deep recess between the waves.

316. _Q._--What other sizes of safety valves are used in locomotives?

_A._--Some are as large as 4 inches diameter, giving 12 square inches of area; and others are as small as 1-3/16 inch diameter, giving 1 square inch of area.

317. _Q._--And are these valves all pressed down by a Salter's spring balance?

_A._--In the great majority of cases they are so, and the lever by which they are pressed down is generally graduated in the proportion of the area of the valve to unity; that is, in the case of a valve of 12 inches area, the long end of the lever to which the spring balance is attached is 12 times the length of the short end, so that the weight or pressure on the balance shows the pressure per square inch on the boiler. In some cases, however, a spiral spring, and in other cases a pile of elliptical springs, is placed directly upon the top of the valve, and it appears desirable that one of the valves at least should be loaded in this manner. It is difficult when the lever is divided in such a proportion as 12 to 1, to get sufficient lift of the valve without a large increase of pressure on the spring; and it appears expedient, therefore, to employ a shorter lever, which involves either a reduction in the area of the valve, or an increased strength in the spring.

318. _Q._--What are the proper dimensions of the steam pa.s.sages?

_A._--In slow working engines the common size of the cylinder pa.s.sages is one twenty-fifth of the area of the cylinder, or one fifth of the diameter of the cylinder, which is the same thing. This proportion corresponds very nearly with one square inch per horse power when the length of the cylinder is about equal to its diameter; and one square inch of area per horse power for the cylinder ports and eduction pa.s.sages answers very well in the case of engines working at the ordinary speed of 220 feet per minute. The area of the steam pipe is usually made less than the area of the eduction pipe, especially when the engine is worked expansively, and with a considerable pressure of steam. In the case of ordinary condensing engines, however, working with the usual pressure of from 4 to 8 lbs. above the atmosphere, the area of the steam pipe is not less than a circular inch per horse power. In such engines the diameter of the steam pipe may be found by the following rule: divide the number of nominal horse power by 0.8 and extract the square root of the quotient, which will be the internal diameter of the steam pipe.