It is customary when a slight crack appears in a boiler plate to repair the leak by drilling small holes at the ends of the crack in order to prevent it from spreading any further, and then caulking the crack and covering it with a patch. This kind of repair can be easily done by any good boiler-maker, but there is one point which should not be overlooked : that it is absolutely necessary that the small holes which are drilled at the ends of the crack be located at the extreme ends of the crack, and not merely near the ends. It is often difficult to tell exactly how far the crack extends, and therefore these holes are sometimes located near the end instead of at the end. In this case, continued use of the boiler will develop the crack beyond the holes, and the trouble must be repaired again.

While ordinary careful investigation may fail to locate the ends of the crack, yet there is a simple way in which this may be safely done. First rub the plate in the vicinity of the crack with oil; wipe the oil off, and cover the plate with chalk. The oil which has penetrated the crack will then be exuded and show plainly the extent of the crack in the chalk, whereupon the holes may be located in their proper places, and either a hard or a soft patch may be applied, according to the position of the damaged plate.

Riveted joints have always been an expensive and troublesome part of a boiler to construct. An ideal boiler shell would be one in which there were no seams, so that the entire shell would be one homogeneous piece of metal, all parts of which were equally strong. A riveted joint must always be weaker than the solid plate unless the plate at the joint is up-set so that its thickness is greater than the thickness of the rest of the shell, an operation which, for practical reasons, is impossible. The best-designed riveted joints rarely have more than 90 per cent. of the strength of the plate which they join, while the ordinary double-riveted lap and butt joints have a very much less efficiency, ranging from 65 to 85 per cent. Furthermore, it takes skilled labour and expensive machinery to lay out the seams, and punch or drill the rivet holes, and drive the rivets, and, even with the best work, there is apt to be trouble in making the joints steamtight, for the rivets must be thoroughly up-set in the holes and the edges of the seams carefully caulked.

The Boiler-maker estimates in the case of a 72-in. by 18-ft. horizontal tubular boiler, built to withstand a pressure of 125 lbs. per square inch, that the total cost of labour and material used in the boiler would be about $561, of which 80 per cent. is the cost of material and 20 per cent. is the cost of labour. Of this amount the cost of labour on the riveted joints alone - that is, of laying out the rivet holes, punching, riveting, etc. - is 19 per cent. of the cost of labour and 3.7 per cent. of the total cost of the boiler. The additional material - that is, butt straps or laps and rivets - amounts to 5 per cent. of the cost of material and 3.8 per cent. of the total cost of the boiler. Therefore the cost of the riveted joints alone is about 8 per cent. of the total cost of the boiler. This is assuming that the longitudinal joint is a double-riveted butt joint, and that the holes are punched and all rivets driven on the machine. If a stronger joint were used, or if the holes were drilled or the rivets driven by hand, the cost of the riveted joints would be increased.

The only possible substitute for riveting seems to be some form of welding in which the metal itself is structurally united in such a manner that the finished product forms one homogeneous piece of uniform quality and properties throughout. Furthermore, in order for such a system to be of any practical use, the tensile strength of the welded joint must be as great as or greater than that of a riveted joint, and the cost of doing the work, including the fixed charges on the apparatus, must not be greater than the cost of riveting; that is, it must be less than 8 or 10 per cent. of the total cost of the boiler.

The ordinary methods of welding by mechanical means, such as hammering, cannot be used in welding boiler shells, both for practical reasons and because the strength of a weld made in this manner is always uncertain.

By the autogenous welding the metal itself is raised to a temperature sufficiently high to cause it to be its own joining material; that is, the parts are joined together by the fusion of their own substance without mechanical aid. Such a process requires that only a small area of the metal in the vicinity of the joint be raised to a high temperature, and for this purpose electricity was first used. While the cost of electricity for this purpose is not excessive, yet it is rarely used, except on castings.

The hydrogen welding is successfully used on plates up to 30 m.m. thickness, and it is claimed that a joint of perhaps 95 per cent. of the strength of the metal can be made. The temperature (2,000° F.) obtainable with the hydrogen blowpipe is, however, somewhat less than the melting point of mild steel (between 2,700 and 2,800° F.), so that for thick plates the heating must be continued for some time, and therefore the consumption of oxygen and hydrogen rapidly increases with the thickness of metal being welded, increasing the cost of welding.

A still more recent development is the use of acetylene-dissous, and this seems to have been fairly successful, since a very high temperature (3,600° F.) can be obtained with it, so that plates up to 20 m.m. thickness can be welded more rapidly.

At present the use of autogenous welding will probably be confined to repair work, for which it seems particularly well adapted on account of its portability. It certainly will not, nor can it be recommended to, be used for welding seams of large boilers or pressure tanks until it is absolutely known that a reliable weld can be made which will be at least 85 per cent. as strong as the metal itself. The results obtained are, besides, largely dependent on the personal skill of the welder both in regard to the quality of the work turned out and the speed at which such welding can be done.