fig. 41, the landing is exposed to the impact of the flame, while joints which are removed from this action by the depth of a furnace corrugation (fig. 42) suffer. On the other hand, it is unreasonable to imagine that the carbon and other impurities of the fuels will not enter the boiler plates if acted upon sufficiently long (see p. 150); at any rate Lowmoor and other good quality irons do not flange well, and are generally not up to their standard if taken out of the furnaces of old boilers. Fatigue and other influences to which boiler plates are exposed may make them brittle, but all these points are fully discussed in the chapter on Strength of Materials.' Here it will only be necessary to mention that the cracks which are sometimes found at the back-end side flanges of furnaces (figs. 43, 44), in double-ended boilers, and in the flanges of Adamson's rings (fig 45), are believed to be due to fatigue-i.e. alternate compression and tension stresses; but other causes, such as inefficient annealing or other bad workmanship, may first have made the plates brittle. FIG. 42 FIG. 43 FIG. 44 FIG. 45 As regards the influence of time, little information is as yet available, but that it produces effects is strongly supported by the experiences published by A. J. Maginnis in Engr., 1885, vol. lx. pp. 447, 475, 504, to the effect that the combustion chamber plates of his boilers cracked after being in use for several years, and that the shells, neck-pieces, and other parts were thoroughly brittle when breaking up the boilers. Here neither fatigue nor gases can have been active. In order to guard against these various troubles care should be taken that all material used for boilers and their repairs should be good-i.e. capable not only of standing certain prescribed tests, but also of resisting the various influences mentioned aboveand the general opinion is that, except as regards tenacity, the very mildest qualities are by far the best. Seams in Furnaces.-Riveted joints should not be exposed to the direct action of the flame. Where this cannot be helped, as in 3 3 3 3 3 FIG. 46 the case of patches, the greatest care should be taken to keep the seams cool. Scale ought not to be allowed to accumulate there, and the two plates should, if possible, be brought into metallic contact, by removing the black scale, by thoroughly washing away the oil used for drilling (for this is one of the worst conductors of heat), and by fitting and bolting the plates so firmly together during riveting that caulking is almost superfluous. It is perhaps due to the more perfect contact between the plates that double-riveted seams behave better than single-riveted ones when exposed to the direction of the flame. Experience has shown that for furnace saddle seams all these precautions are unavailing under the action of forced draught, and it has been found that when they must be repaired in this way, instead of using countersunk rivets, very large snap-headed ones should be fitted (fig. 46), with the object of reducing the exposed surface of the outer landing as much as possible, and at the same time carrying off all the heat through the rivets into the water. (See p. 119.) Seaton quotes Wye Williams, N.A.,' 1894, v. 35, p. 284, as having fitted 6inch studs through his furnace plates, and found that the fire ends burnt off to a length of 2 inches, so that the much shorter rivet heads can take no harm. Whenever possible, patches should be fitted on the fire sides of defective plates, so that when the rivet holes crack, as shown in fig. FIG. 47 47, it will not be necessary to cut out more material and fit a larger patch, as shown in dotted lines, when a renewal is necessary. Cracked Rivet Holes.-One of the chief causes of these rivet holes cracking is scale, and whenever possible it should be removed from these patches. In the chapter on Heat Transmission' explanations will be found why the general efficiency of a boiler is not seriously affected by accumulations of scale, but that this substance will cause a considerable rise in the temperature of the plates it covers, which leads, as has just been explained, to cracking or to other troubles, such as leaky tubes and collapsed flues. The idea that scale prevents corrosion may still prevail, and may even be a true one; but the more rational view, that there are other and better means of stopping this waste, is gaining ground, and evaporators, for the supply of distilled water, are coming more and more into use. The removal of one evil often produces another, and recently much damage has been done when using fresh water by oil deposits, which, like all scales and other non-conductors, effect a considerable rise in the temperature of the plates they cover. This necessitates the addition of oil filters, which are inserted between the pumps and the boilers. A description of one of these, together with numerous facts and analyses of boiler deposits, will be found in a paper read by Mr. Edminston, N. E. C. I.,' 1892, vol. viii. Recently the design of these filters has been much modified, and very favourable results are being obtained by means of appliances which separate the grease from the steam. Boiler Scale.-Sea-water should not get into a marine boiler, but when it does enter, either by accident or intentionally to form a protecting scale, it meets with grease and iron rust, its own production, and the result is scale and deposit. Professor V. Lewis, 'N.A.,' 1891, v. 32, p. 67, deals with such a case, and gives analyses of scale from various parts; these have been recalculated in terms of sulphate of limecalcic sulphate. scale depends Thus, for some It will at once be seen that the composition of the largely on the position from which it has been taken. reason or other, calcic carbonate is not found in the deposits on the boiler bottom, while grease is found chiefly in the deposits on the tubes, showing that it has a buoyant effect on the constituents to which it attaches itself. It seems to have most affinity for magnesic hydrate and also much affinity for iron rust. When the latter predominates the compound is so heavy that it sinks to the boiler bottom; when magnesia is in excess a deposit is formed on the tubes. Magnesic hydrate, as is well known, is the very dangerous deposit known as 'floury deposit,' and is generally due to an excess of alkalies, which, as will be seen, seem to accompany the magnesia in a fairly constant proportion. The buoyant action of the grease shows itself in the scale on the furnace bottom and on the tubes. It is strange that there is so little grease on the furnace top-possibly heat has driven it away; but then, one would expect to find an excess of magnesia. The Effect of Scale. It is generally believed that scale causes collapses and bulges simply because it is a bad heat conductor; but a recent experience (M. S. U. A.,' 1899, p. 6) throws some unexpected light on this subject. An accident is there described in which a furnace collapsed and rent although the scale was intact over the bulged part. Near the rent there was another bulge which could only be seen from the fire side; on the water side it was covered by an intact sheet of scale about 3 in. thick. The hollow space between the scale and the bulged plate must, while the boiler was at work, have been filled with superheated steam, which of course is an excellent non-conductor of heat. When once the plate separated from the scale the hollow thus formed would effectually shield the plate from the cooling contact with water, and it would grow red hot and bulge. This experience suggests that hard scale is more dangerous than soft scale, also that irregular working of boilers, which, on account of changes of temperature, would cause the hard scale to separate from the iron plate, should be avoided; and further, that a boiler which has once been emptied of water should not be refilled before at least all loose scale has been removed. The suggestion that hard scale may be more dangerous than soft scale derives additional support from another accident to a boiler, whose furnace crowns were covered 2 inches deep with a thick muddy deposit; the sides were of course bare, yet it was the sides which collapsed when the water grew too dense through salt. Sea Water. A most elaborate series of chemical analyses of sea water has been carried out by Prof. W. Dittmar for the Challenger' Report; see article Sea Water, Encycl. Brit.,' p. 611, from which it appears that sea water consists of Chloride of Sodium Chloride of Magnesium Sulphate of Magnesia Sulphate of Lime. Sulphate of Potash Bromide of Magnesium Carbonate of Lime Total Sulphate of lime is practically the only constituent of sea water which goes to form boiler scale, as can be seen by referring to the several analyses mentioned above. Oxide of magnesium is present in small quantities, say from 5 to 10 % of the sulphate of lime. The moisture in the scale should of course not be counted, nor the iron rust, which is due to the wasting of the boiler material. It is thus evident that boiler water contains nearly all the sea water salts except the sulphate of lime. These salts are: common salt, chloride of potash and sulphate and chloride of magnesia. The weights of these salts per 1,000 parts of water will be approximately as follows: Chloride and Bromide of Sodium and of Potassium about Magnesium Sulphate Total 36.0 5.0 2.2 1.1 44.3 In former days these soluble salts, which were of course being concentrated in the boiler, often led to overheating, and even now with leaky condensers or negligent feeding such accidents sometimes occur. Engineers are generally instructed not to allow their boiler water to exceed, which is about four times the density of sea water; but salt water does not commence to crystallise until the density when cold exceeds 1.2045 or about 3. As will be seen from the following table, the amount of salt dissolved by water increases with the temperature, and at 350° F.-a not uncommon one in modern boilersthe weight of salt in water probably exceeds 57 ounces per gallon, whereas when cold it falls to about 50 ounces. The following table has been calculated from Dr. H. Landholt and Dr. R. Bornstein's tables. Salt Deposits. The beautiful salt crystals which are found in boilers, in which the density was so great as to cause collapse, are not the cause of the accident; they have been formed during the period of cooling. The real injury has been done by the dense salt scale, which contains only fine crystals, or still more probably by a local deposit of salt which may have re-dissolved. One case is known in which the furnaces did not collapse nor showed any signs of weakness, although covered with large salt crystals about in. cube. The |