Before attempting such repairs, careful measurements should be taken, and, with the help of a board, representing the furnace front,

and a wooden model of the furnace (fig 64), the action of passing the latter through the former should be performed. This will also enable

one to determine how much of the back end has to be cut away. Furnace Patches.-Furnaces sometimes suffer severely from pitting along the line of fire bars. If some of the pit holes should have pierced the plate, it is necessary to

drill them out and fit screwed studs, or, better still, broadheaded rivets, which may be square, oval, or irregular in section, to suit the special case. Neither give trouble by leakage, but they do not add to the strength of the furnace, and, unless absolutely necessary, it is better not to perforate it at any point above the bars. If the pitting is uniformly distributed, as well as deep, there is danger of this part of the furnace giving way, but as yet no such case seems to have happened. To guard against this, or to prevent further pitting, some engineers bolt doubling plates, well bedded in red-lead cement, along the line of fire bars (fig. 65). It is difficult to imagine a more inappropriate remedy, for to place here, at the hottest part of the furnace, not only two thicknesses of plate, but actually to separate them by a highly non-conducting material, is little better than inviting disaster. If any action is necessary, the furnace crowns should be removed.

FIG. 64

FIG. 65

Furnace bottoms are sometimes doubled in the same way; but this is not a permanent repair, and the plates should occasionally be removed for examination (see p. 43).

Blisters in the furnace crowns need only be pared away, unless they are so large and deep that there is fear of their weakening the furnace. In that case they have to be cut out and a patch fitted.

Cracked Furnaces.-Plain furnaces have hardly ever been known to crack except at the flanges or the seams, but every one of the patented forms has done so. If due to scale or grease, the cracks start on the fire side and sometimes run longitudinally; if due to unequal expansion or excessive strains, they sometimes show first on the water side, and generally in a circumferential direction. Similar cracks (see fig. 43, p. 26) are sometimes found in the sides of the furnace saddles, and can often be repaired by chain-pinning them. They make their appearance on the fire side, and sometimes, if not interfered with, do not penetrate through the plate. It is difficult to say whether they are due to excessive strains or to injured material, such as working the flanges at a blue heat, or to both causes. The manufacturers take great pains to thoroughly anneal the patent flues before sending them out, but unfortunately boiler-makers find it more convenient to reheat the corners while fitting them to the tube plates than to alter the latter to the required shape. In some works these corners are even welded, and possibly not annealed afterwards. But the greatest danger seems to be incurred if heaters are applied to these points. It has been suggested that the furnaces should be heavily hammered after the hydraulic test.

When the cracks, which show themselves, are so serious that tapping pins into them would not prevent leakage, the affected parts have to be cut out and patches fitted. In fig. 66 the patch is fitted on the fire side, and can easily be renewed if the rivet holes should crack. In fig. 67 it is fitted on the water side. This is done with

the object of keeping the seam turned away from the fire, but it is a most difficult piece of work, because the seams at A and B have to be sprung open. This patch is also not to be recommended, because if the rivet holes should crack, a larger corner would now have to be cut out and replaced (see fig. 47, p. 26). If the tube plate or furnace saddle is not too stiff, the repair shown in fig 67a is a very convenient one.

Combustion Chamber Bottoms.-These are often very seriously corroded on the fire side, particularly near leaky seams, and, as even the rivet heads retain their shape when wasted (see fig. 68), there is danger that such defects will be overlooked.

FIG. 68

The salt from leaky tubes when mixed with ashes causes corrosion at these points, and to prevent this, the bottoms are sometimes cemented. This is a very dangerous practice; not only does this prevent any leakage from being detected, but if there is one, or even if moisture finds its way between the plates and the cement, the furnace heat will convert it into steam and cause a small explosion, which may do damage to bridges and fire doors, &c. Fitting doubling plates to these parts may lead to the same result, but the consequences would be still more serious.

The corrosion on the water side of the combustion chamber is often very evenly distributed, and if the bottoms are found to be thin, the backs are most likely worn away too, and before starting to cut out the one it is best to drill holes in the other for measuring the thickness.

The best means for detecting weak places in plates is to make a thorough examination when the plates are well scaled. The shadows produced by the lamp often indicate irregularities, particularly pit holes. The hammer will give no reliable indication of thickness when this exceeds in., and then it is of course possible to dent the plate and even to knock a hole into it. When general wasting and consequently structural weakness are suspected, drilling the plates is the only means of measuring their thickness, though magnetic and electric tests have been suggested and might be made available.

Before cutting out thin plates, it is well to drill test holes beyond the intended new seam, for if further thin places are revealed after the plate has been cut across, it will be necessary to repeat the operation. Patches on the lower parts of the combustion chamber usually include flanges, and care should be taken that these are properly fitted, for none of the seams or rivets can be caulked from the water side, as is the case with new boilers. Of course new stays will have to be fitted, and these should always be nutted, no matter whether this was originally the case or not.

Screw Stays very often lose their nuts by burning, or rather bad welds open under the influence of the heat (fig. 69). It is quite a common practice to cut off the greater part of the projecting stay and rivet it over. not be done, because it is due to the presence of the

FIG. 69

This should nuts that a

reduction is allowed in the thickness of the plates. The better plan is to cut the thread deeper (fig. 70), for which special tools exist, and to fit new nuts.

This burning away of the stay nuts is generally due to their not being in metallic contact either with the stay or the plate; or, in other words, they are a loose fit on the thread, and are resting on a washer and two thicknesses of red-lead cement. When refitting the nuts, these mistakes should not be repeated, but where the stays are not normal to the inside plates, taper washers are a necessary evil. This arrangement should be objected to in new boilers.

With thin plates the stay ends often leak, particularly if their ends are riveted over. Instead of renewing them, some boiler-makers go to the trouble of making a small cap and bolting it over the head, as shown in fig. 71. Sometimes several stays in one combustion chamber are hidden away in this manner, and it is only a wonder that this practice has never led to a disaster. (See p. 226.)

The excuse made for fitting these patches is that the plates had become too thin for holding a new stay. If the thinning is only local, a very efficient repair is to replace the stay by one of a larger diameter. But in screwing up the nut its power will perhaps be so great that the thread in the plate gets stripped; to guard against this, a check nut should be fitted on the inside. It is bad for the transmission of heat, but is better than no support at all.

Often it will be necessary to remove the stay, and to fit a small patch and tap it, as shown in fig. 72. This is the best possible arrangement, particularly if the stay is not normal to the plate, because the patch could be made taper, bent or recessed so as to let the nut come into metallic contact with it.

Sometimes the plates remain uninjured while the screw stays waste away and will have to be renewed, generally of a larger diameter, the threads having been injured.

In order to effect these various repairs, and to stop external leakages, sufficient space should be left at the backs of boilers, so that a man could work there.

It will be noticed that those stays which are nearest the combus

tion chamber bottoms and sides give the most trouble. This is no doubt due to the very severe strains to which they are subjected through the unequal expansion of furnace and shell plate, and therefore they ought to be kept as far away from the flanges as the strength of the boiler back plate will allow (see p. 140).

Similar remarks apply to the top row of screw stays to boiler shells; if placed too near the combustion chamber top they also are subjected to very severe strains, sometimes leading to rupture.

Tube-Plate Troubles.The last of the troubles in combustion chambers is the tube plate. An explanation as to the causes of leaky tubes has already been attempted, and some remedies suggested. The repeated re-expanding, together with the injury caused to the metal by

оо

оо

FIG. 73 a

оо

ооо

FIG. 73 b

getting hot, if covered with scale, ultimately causes the tube plate to crack. Fortunately, since steel and thicker tube plates have been introduced, such cracks are rarely met with. If caulking will not stop the leak, or if several spacings show cracks, it is necessary to bolt little spectacle patches (fig. 73a) over them, preferably on both sides. Keys, shaped like a dumb-bell (fig. 736), are said to have given good results, but the fitting must be carefully done to make it a good job.

If the front tube plates of boilers are curved or bent back at the top, they are subjected to very serious tension stresses, which in one case led to a rupture, so that it was found necessary to fit vertical stays, as shown in fig. 74. It is, however, questionable whether the stays were

of much value, as their lower extremities had only been bolted to the furnace tops, which, being thin, could not have offered a resistance if the large stays had been strained to their utmost.

Repairs to the Boiler Shells are sometimes necessary on account of the wasting away or cracking of circumferential seams. As such patches are not required for strength, but only for water-tightness, they need not be thick. At one time it was customary to make them of cast brass, a thick sheet of lead, which was hammered over the seam, being used as a pattern; but it was found that this gave rise to very serious