partly to the wire at V. The rocking frames (fig. 429) are made of thin steel plates, sharpened at the bottom. The wire passes over the notch y. By fixing several wires round one boiler the straining of the various parts of the shell and the influence of the end plates and seams can easily be ascertained (see p. 199).

Most of the other parts of the boiler can best be measured by rods or battens having a micrometer screw or other contrivance attached

to their ends. Fig. 430 shows a convenient instrument for this purpose: S is one of a large number of small sleeves, to which ordinary wood screws have been brazed, and by which means they are secured to the wooden rod or batten; R R is a steel rod which easily fits the sleeve S, and is accurately graduated, preferably in millimetres; V is another sleeve, slit open on one side, and containing a vernier. The two should be a sliding fit. Measurements of the various points of a boiler before and after testing are taken by consecutively inserting the rod R R with its vernier into all the sleeves S and noting the readings (see p. 175).

For the measurement of furnace deflections, ordinary trammels are more convenient, as, by their means, it is easier to detect the major and minor axes of deformation.

For instance, if measurements are taken at the axes A A and B B

(fig. 431), with a deformation as shown by the dotted lines, then the final readings would differ but little from the original ones. However, if trammels are used, they will show that the diameters had changed their angle, and the trammel resting at A would touch e instead of a, while the one resting at B would touch f instead of d. This would lead to further measurements being taken at jh and ik, and would show that a serious deformation was taking place there. Should one

of the diameters be so much reduced that the trammel cannot be got into position, its points of touching must be marked along the axis of the flue.

In order to reduce these measurements to absolute readings, employ the following formula: A, and A2 are the difference between the diameter of the furnace and the length, D, of the trammel; x and y are shown in fig. 432.

If the original reading was x and the subsequent one y, then the

deformation of the furnace was 4, +42

=

Let D 40 ins., x = 4 ins., y = 2 ins.,

x2 + y2
8D

then

If the first measurement of x was 6 ins., and the second one 4 ins.,

then the deformation was also 16 in., viz.

36-16
8 × 40°

Having tested the boiler, it ought to be examined internally, after which it is usual to cement the lower seams of the shell, to cut the various openings for steam pipes, &c., and to fit the boiler on board.

On account of the numerous recent cracks in furnace saddles, it has been suggested that these parts should be severely hammered after the test.

308

CHAPTER IX

DESIGN

The Proportions of Heating and Grate Surfaces and sectional areas of tubes, funnel, and other parts of boilers vary considerably according to the experiences of the manufacturers or shipowners, and it would be rash to attempt to harmonise these divergent views; all that can be attempted is to reduce the problem to the simplest elements, and these, it is hoped, will indicate how to make comparisons and draw conclusions from authentic records. The two conflicting factors are economy on the one hand, and maximum performance for a given weight of boiler on the other. In the one case large heating surfaces are necessary, in the other a high temperature of the escaping products is unavoidable.

Assuming the case of a boiler supplied with feed water of 100° F. and worked at a pressure of 150 lbs. (water temperature 315° F.), and assuming that the fuel is capable of evaporating 15 lbs. of water at and from 212° F., equal to 12.3 lbs. under the above conditions, and assuming that 15 lbs. of steam are required for each I.H.P., then the following results may be expected :

With the help of the above-mentioned funnel temperatures it is possible to estimate the force of its draught when the height is known.

Funnel Draught Measured in Inches, Water.

Any desired pressure or suction can of course be obtained by mechanical means (forced draught). If the specific resistance due to fuel and other obstructions were known, it would be easy to estimate Q, the weight of air which passes through 1 square foot of grate per hour: Q = 730 where h is the draught pressure

1+

h

in inches of water, and r the specific resistance to the air passage; but as this latter value can only be guessed at (see p. 101), and as the weight of coal consumed is not strictly proportional to the weight of air which passes through the grate, the matter is simplified by comparing the coal consumption and funnel draught direct.

Approximate Coal Consumption in lbs. per Square Foot of Grate. Draught pressure, inches (water)

1 2 3 4

20

888

Pounds of coal burnt per hour per sq. from 15 20 25 30 40 50 60 ft. of grate

to 20 25 30 40 50 75 80

The following case will illustrate the use to which these tables can be put. The relations between the various surfaces are to be determined for a boiler whose funnel height is 80 feet, and which is to give economical results. The uptake temperature should therefore not exceed 500° F., and the draught suction will be about 6 in.; the consumption per square foot of grate will be about 20 lbs. per hour; the indicated horse-power will be about 11 per square foot of grate, and the heating surface will have to be about 55 times as large as the grate area.

If economy is of secondary consideration, then the uptake temperature will have to be about 1,000°, the draught will be 85 in., the consumption 30 lbs. per square foot of grate, and the horse-power about 14, which is not much more than in the above case; but the heating surface will be very much less, viz. 23 times as large as the grate.

Had the funnel been shorter, the consumption per square foot of grate would have been less, and the ratio of the heating surface would also have been reduced, whereas with forced draught it would have to be increased.

The absence of any reliable data on the relationship between funnel height, grate area, and heating surface makes it difficult to check the above values by actual performances, and the above tables and calculations should, therefore, only be looked upon as an indication how deductions could be drawn from really reliable information. The necessity for doing this has occasionally made itself felt in boilers which were to be exceptionally economical. In one case the lowness of the funnel and its temperature reduced the boiler performance so seriously that it was necessary to reduce the heating surface by blocking up a large number of tubes; and it is evident that if these boilers had originally been properly designed, they could have been made of very much smaller dimensions.

Funnel Dimensions.-A natural desire to reduce both the height and diameter of a funnel, so as to offer little resistance to the speed of the ship, will, if carried too far, reduce the draught, and with it the

boiler performance. Many people hold the view that a short funnel with a large diameter is as efficient as a tall one with a reduced sectional area; but this can only be true in cases where the resistance to the motion of the products of combustion is greatest in the funnel. In well-designed ones this is far from being the case, and then, within certain limits, the draught is not affected by the diameter. The one limit is determined by the velocity of the waste gases, which should not exceed 25 ft. per second under natural draught. The other limit is more difficult to fix, but it is quite certain that if a funnel is made too large in section, cold air will rush down from above and interfere with the up current. This happens with those funnels which occasionally draw well, and at other times badly. It is well known that factory chimneys parallel outside and conical inside, which were the fashion some years ago, were serious offenders in this respect, from which it is reasonable to conclude that not only the size and proportions, but also the shapes of funnels, affect the results.

Dimensions of Grates and Furnaces. As a general rule, and for ordinary work, the stoking of furnaces whose bars are longer than 5 ft. cannot be done economically. When their lengths exceed 6 ft. most of the extra coal burnt is simply wasted. This is particularly the case with forced draught, and some engineers advocate that under such conditions the grate should not be longer than 4 ft.

Stoking is seriously interfered with if the furnace diameters are small, and probably the cooling influence of the plates close to the fire retards combustion, whereby unconsumed gases are permitted to escape. So that, unless grates are very short, furnaces should not be made less than 3 ft. diameter, and where flaming coals are used they should be still larger.

Builders prefer small furnaces, because with them more grate surface and more heating surface can be got into a boiler of a given diameter; but steam users should see that this does not lead to their being supplied with an inferior boiler.

Tubes. The furnace and tube lengths are practically identical, and vary from 4 ft. for short double-ended boilers to 9 ft. for single-ended boilers with forced draught. In the latter case the tube diameters are about 2 ins., while for natural draught, where a small internal sectional area would offer too much resistance, the length is usually about 24 diameters.

Tube Surfaces, and Space Occupied. The following table gives some of the dimensions of boiler tubes. The thickness of the metal is assumed to be 15 in., or about No. 9 wire gauge. The external and internal heating surfaces are given per foot of length. The internal sectional area of a single tube is also added. The last two lines of the table contain the amount of end space required by one tube when it is surrounded either by 1 in. or 1 in. of water space (see dotted lines, fig. 433).

о

FIG. 433

In one and the same boiler the number of tubes for each furnace sometimes differs considerably. This is a bad design, for there are either too few for the one or too many for the other. The tube surface amounts to from 75 to 85 % of the total heating surface.