It has also been shown (see p. 222) that if the rows of rivets are placed very wide apart, the stresses are not uniformly distributed amongst them. This tends to lower the value of an ordinary joint, but the unequal straining would be somewhat reduced by cutting away the metal along the dotted line (fig. 444). In the absence of any conclusive experiments or investigations, it is safest to adhere to the general practice mentioned above. In double-ended boilers it will often be found that the percentage of plate remaining between the holes of the screwed stays is less than that between the rivets of the longitudinal seams. The stays will then have to be pitched diagonally. But some allowance might be made for the absence of bearing pressure (see p. 216).

Lightest Joints. For the purpose of comparing the weight of various joints, and also the amount of metal removed by drilling, which is a sort of measure of the labour expended in the shops, the above drawings of riveted joints have been made; in every case the holes are placed at an angle of 60°. The dimensions are in ac

cordance with Lloyd's Boiler Rule for a steel plate 1 in. thick, the rivet diameters not being less than 1 in., and the percentage of the plate and of the rivet sections being equal. The estimates are contained in the following table :

The last column contains the quotients obtained by dividing the total weight of 12 ins. of length of the boiler shell by the respective working pressures, as found by the

rules; these are convenient measures of the efficiencies of the various joints. It will be seen that the quadruple lap joint (fig. 446), though simpler, is almost as good as the double-riveted butt joint (fig. 443), and that the common buttstrap joint (fig. 442) is nearly equal to the one shown in fig. 444. This order is slightly changed if the circumference

FIG. 445

of the shell is made up of only one or two instead of three plates. Of course, in works where the machines are incapable of

Y

dealing with plates of more than a certain thickness it may be necessary to use joints of the very highest percentage, and then the one sketched in fig. 444 is undoubtedly the best (see p. 222).

FIG. 446

Best Arrangement of Staying Flat Plates.-A comparison similar to the above can be carried out for flat plates. Here it cannot be a

FIG. 447

question of variation of pitch, for it is quite clear that the total weight of stays is a constant quantity, while the weight of the flat plates

FIG. 448

varies as the thickness, or inversely as the pitch of the stays, and by placing these very close together the plates could be made quite thin.

In the following table the pitch of the stays in the combustion chamber plates is 10 ins. each way, the working pressure is 100 lbs., and the diameter of the screwed stay 1 in.

In the following table the area of steam-space plating supported by each stay of 21 ins. diameter is 300 square ins.

Here, again, it will be seen that the mode of attachment which permits of the use of the thinnest plate does not lead to the lightest construction.

Doubling strips and plates are sometimes arranged as shown in fig. 449, the adjoining plates overlapping each other. In such cases the corners should be cut away, so as to reduce the length of the taper.

In the case illustrated in fig. 450, the plate has evidently been weakened by the seam. Doubling plates are also fitted to the front tube plate between the nests of tubes, and to the back plates of the boilers wherever the stays are too wide apart, and here, too, the aps are sometimes made very wide. Of course, the number or the sectional area of the adjoining stays or stay tubes must be increased to bear the extra load. Angle irons are also fitted, but usually in such a manner that they give no support.

These few suggestions about the design of certain parts of boilers, so that they shall be as light as possible, can readily be extended to

other parts, and also to the problem of cheapness and despatch; but when the main principles have been settled, care must still be exercised in carrying them out, for other circumstances may come into play. This occurs in the following simple design of a steel boiler in accordance with Lloyd's Register Boiler Rules. It is intended to illustrate how the various rules can be applied, and the various references to the tables will show how to use them advantageously.

In order to shorten the explanations, practically no notice has been taken of mechanical difficulties, which are discussed elsewhere, and the design is necessarily imperfect in these respects.

The External Diameter of the shell is not to exceed 13 ft.=156 ins., the heating surface is to be 2,000 square ft., and the grate surface 50 square ft., but the length of the grate is to be limited to 5 ft. All the vertical water spaces are to be 10 ins. wide, and the steam-space stays are not to be placed closer together than 16 ins. The water space round each tube is to be 1 in., round the furnaces 4 ins., and above them 7 ins. Working pressure, 160 lbs.

The Furnace Diameter is found by dividing 50 square ft. by 5 ft., which gives 114 ins. of furnace fronts. Two furnaces would be too large, but three of 38 ins. internal and about 39 ins. external diameter will be convenient (see pp. 101, 313, and table, p. 346).

311.

The Tube End Spaces have to be estimated, as explained on p. The internal diameter of the shell is about 154 ins., and the water spaces, as explained on p. 311, are 9 ins. at the wings, 9 ins. between the nest of tubes, and 7 ins. above the furnaces. The steam space has been made one-third of the boiler diameter.

Tubes. If the tubes are made 31 ins. diameter, then each one will occupy an end space of (41)2 = 181 square ins. Dividing this into 4,330 gives 240 as the necessary number of tubes. The tube surface will be about 80 per cent. of the total-say, 1,600 square ft.-and therefore each tube must have a surface of 67 square ft., which necessitates that its length should be 7.9 ft. (see table, p. 311). If 3-in. or 31-in. tubes had been decided upon, these numbers would have been respectively 272 and 7.5 ft. and 215 and 8.1 ft. These 240 tubes should be arranged in bundles of as nearly as possible equal numbers (78), and if possible in such a manner that the number of vertical and horizontal rows are all odd numbers, for then the staying is very much simplified.

The circular line A B (fig. 451) should be drawn; it marks off the boundary for the centres of the extreme tubes. Its radius is 663 ins. Measurement or calculation will show that after deducting the central water spaces there remain 121 ins. for the horizontal spacing of the tubes, which, as they are placed 41 ins. apart, number 28. The lines CD and E F, 7 ins. above the furnace crowns, indicate the lower boundary for the tube spaces, while N M is the upper limit. The height between these lines is 34 at the wings and 52 ins. at the centre, equal to about 8 and 12 tube pitches. The tube nests can now be arranged as shown in fig. 451, viz. 8 × 10 + 12 × 7 + 8 × 10 = 244 (odd numbers of rows being impossible). Of these, 6 fall away at the four corners, leaving 3 more than required. This agreement is sufficiently close for all practical purposes, but before proceeding it is advisable to estimate the

=

Heating Surface in the Combustion Chambers, so as to be sure that the total of 2,000 square ft. is reached. The clear depth, according to p. 313, is 31 ins., and the heating surface other than that of the tubes will be about 380 square ft., which, added to 238 × 6·8: 1,620, is exactly 2,000 square ft. Had this result not been obtained, then the tube lengths would have had to be altered a little. The boiler length is 11 ft. 3 ins.

The Lengths of the Shell Plates can now be fixed, and the riveted joints arranged.