Imágenes de páginas


these parts, quickly loosen all rust as it is formed, so that metallic iron or steel is always exposed to the air if allowed to be produced there. Certain it is that if there is any pitting going on in a boiler,

the greater part is sure to be found along the line of fire bars, Another part which is also severely attacked is the under side of the furnace and combustion chamber, for here the air bubbles cannot rise if they have once been formed.

In support of these views it may be pointed out that severe pitting frequently takes place on the forward side of propeller blades at the leading edge, as shown in fig. 95. Here the air is not driven out by heat, but is abstracted by the partial vacuum which is found there, and, in spite of the high velocity of the water, there seems to be sufficient time for the mischief to be done. Even bronze blades are sometimes pitted at this point.

Another confirmation is found in the severe pitting of internal iron feed pipes ; even copper ones waste away. Here the air is liberated by

the transmitted heat of the surrounding water, Fig. 95

and it has been suggested that long internal iron feed pipes should be fitted to boilers, and renewed whenever they are eaten through, for, whether this is due to the air or some other corrosive agent, it is cheaper to lose a regular quantity of temporary piping than to have to renew furnaces. They last about eighteen months.

Distribution of Corrosion.--One peculiarity about corrosion in general is that boilers which suffer much at their lower parts are often found to be as good as new in their steam spaces. In others the stays and plates of the steam space suffer severely, while the lower parts are only slightly attacked. This happens chiefly when the feed is discharged near the water level, but, as it is next to impossible to say in what path the circulation in a boiler takes place, and whether the feed is carried up or down, definite conclusions cannot be drawn. Another curious fact is, that if a number of boilers are connected with a single feed pipe, that one which is farthest away from the pumps suffers more than the others.

Steam-Space Corrosion. If it is the air which attacks the steamspace stays and superheater plates, there is perhaps no other remedy than not to admit it, unless it can be proved that zinc here also acts as a protector. It seems to do so, but this can hardly be through galvanic action. Possibly here again it is the presence of zinc salts, of which it is not difficult to imagine that they have been carried into the steam space. In some steamers these parts are whitewashed with zinc white, and the result seems to be a satisfactory one.

It is of course possible that this corrosion is due to steam alone, or to the hydrochloric acid which, as has been mentioned, escapes from sea water when evaporated in iron boilers; but it is safer to blame the air, as it has not yet been shown that steam can attack iron at temperatures ranging near the boiling point of water. If it did, the action ought to be equally strong in all boilers, and that is far from being

the case.

That this form of corrosion may lead to serious results is only to be expected, because of its rare occurrence and on account of the wasting being very uniform. Thus, rivet heads and flat plates in steam spaces retain their original shape even when most of their substance is gone. As an instance a case may be mentioned where the rivet heads of a steam dome were corroded as shown (fig. 96). The head seemed to be intact, but had in reality disappeared, and several of the rivets could be driven back with a hand hammer. The dotted line shows the original thickness of the plate and size of rivet head.

In another case the steam-space end plate of a boiler was wasted as shown (fig. 97), and it was only due to the necessity of renewing the stays that the corrosion of the plate was discovered, for the shoulder round the nut had been mistaken for the washer which is

[blocks in formation]

sometimes placed there. The wasting of the front plates and front ends of the steam stays is aggravated by the heat of the uptake; but, as equally bad cases are met with when baffle plates are fitted, the temperature cannot be the sole cause.

The following is a curious illustration of this sort of corrosion. On examining a pair of boilers of a ship which was barely four years old, it was found that, although efficient baffle plates had been fitted, five of the steam-space stays, situated about 2 ft. above the water level, had lost nearly 30 % of their substance, but only close to the smoke-box end. The starboard boiler was in a perfect condition. Zinc had been used in both, but had been given up. There was only one difference in the structure of these boilers, which is so small that it would not be worth mentioning were it not in connection with the feed. It was found that the knee pipes, K (fig. 98), which had been screwed to the feed inlets, in order to produce downward currents, had fallen off in the port boiler, and therefore a possibility exists that the feed water of that boiler travelled in the direction shown by the arrow F, and would part with its air sooner and deliver it into the steam space.

Internal Feed Pipes burst, although their ends are open ; this is no doubt due to water-hammer action. If the boiler water contains salt, its boiling point will be higher (see table, p. 60) than that of the

pure feed temporarily stopped ; there is every probability that the pipe will have contained steam. As soon as the feed supply was

restarted the steam condensed, and the boiler water was sucked towards the valve with sufficient velocity to injure the joints or burst the pipe. A small hole near the flange or an uncemented joint entirely prevents this action.

A most salutary lesson was taught some years ago, when the admixture of air to the boiler feed was advocated as an improvement to the circulation. Its injurious effect was so great that the idea very soon died out. The Boiler Committee also helped to convince

engineers on this point, and FIG. 98

in all high-class engines

care is taken to keep the air out of the boiler, this being the simplest and safest plan for guarding against its corrosive action.

Separate automatic feed pumps are the only efficient means of securing this object.

Galvanic Action.-The laws which have been discovered in this branch of science may be briefly stated as follows: A continuous electric current is only possible in a circuit. Its intensity, measured in ampères, is equal at all points of this circuit, and is proportional to the algebraical sum of the electromotive forces in the circuit (measured in volts), and inversely proportional to the sum of the electrical resistances of the circuit (measured in ohms).

The electromotive force makes its appearance under various circumstances, viz. when two different substances are brought into contact, or when changes of electricity or magnetism take place near the circuit, or when heat is made to travel along part of a circuit.

a Here only the former case will be dealt with.

If a circuit is constructed consisting only of solids, then no current will be generated, because the sum of electromotive forces equals nought. Therefore, if the electromotive force due to contact of one metal with several others is known, that between two of these is found by the difference. The following example will illustrate this.

The electromotive forces in volts due to contact with iron are

[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors]

To find the value when, for instance, carbon and copper are brought into contact we have -485 (-146) = 339. These values change considerably with rises of temperature. If, therefore, one of the points of contact in a metallic circuit is heated, the electromotive force at this point will be altered, and a current produced. Thus, at 530° F. the electromotive force between copper and iron is reduced to 0, and a circuit consisting of only these two metals, but with one joint heated and the other cold, would show an excess of potential of - •146 volt in one direction. All thermo-electric piles are constructed on this principle. With Auids, or with solids and fluids, this simple law does not exist, as will be seen from the following values of electromotive force which appear on immersing any of the above solids in pure distilled water, and also in sea water.

Table.Electromotive Forces

[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]

The electromotive force in the circuit (fig. 99) would therefore be as follows: Contact copper and iron.

- 146 volt
iron and salt water - .605
salt water and copper

+ 475
Electromotive force in circuit .276 volt


A Daniell's cell, consisting of copper, zinc, sulphate of zinc, porous cell, sulphate of copper, has an electro- Fig. 99 motive force of about 1.10 volt.

One effect of a current, circulating in a circuit, is to generate heat at every point, the amount being proportional to the square of its intensity and to the local resistance.

Thus, if the entire resistance of a circuit is R ohms, its electromotive force E volts, and its intensity I ampères, we have


E2 and W=I?. R=

= E. I. R'

R Here W is the amount of work done in the whole circuit measured in watts, of which 746 are equal to one horse-power and 1,049 for one second are equal to 1 heat unit.

Another effect of the electric current is to produce chemical changes in the fluids through which it passes, usually splitting them up into their elements. It has been found that the amounts are strictly proportional to their atomic weights. Thus a current of one ampère, passing through water during one second, will produce •0001038 gram of hydrogen (H2), and 0:5 x 15.96 times as much oxygen (O), viz. •0008283 gram. If the electricity was produced by a Daniell's cell, which then of course forms part of the circuit, it would be found that during this second its zinc has lost 003367 gram, and its copper gained .003279 gram, these quantities being proportional to their respective atomic weights. Iron would have lost .002900 gram.


This loss of metal at the electrodes is said to be a secondary action, being due to the dissolving power of the elements produced there. Oxygen and hydrogen would be liberated from salt water; but the former gas, being nascent, would combine with the iron, if that is the metal of the electrodes.

It is evident that the chemicals produced by these secondary actions must influence the electromotive forces in the circuit; this is called polarisation. Iron shows this property in a very marked degree.

If the positive electrode is made of iron, which, as has been stated, is a distinctly electropositive metal, it will very soon change its nature, and grow even more electronegative than copper. The unexposed parts will still be electropositive. If the whole piece of the iron is now placed in the fluid, a strong current is set up from its negative to its positive end, and through the fluid back to the negative part; but very soon the current ceases, and on examination it will be found that the electronegative property has spread over the whole piece. There are other means for making iron electronegative, viz. dipping it into concentrated nitric acid, or heating it in air-in fact, anything that will oxidise it; and there seems little doubt but that this form of polarisation is due to a scale of oxide of iron, which in some cases is so fine as to be invisible. This would lead to the conclusion that iron ought not to rust beyond its initial stage ; but, as it does so, particularly in boilers, there is no alternative but to admit that this beneficial change is not possible at a boiling temperature, or that the galvanic action, which produces it and causes it to spread, does not exist in a boiler, or will not act in the same way on large surfaces as on small ones. Electronegative iron can be brought back to its primary condition by making it the negative electrode. This is probably due to the reducing action of the hydrogen on the oxide of iron.

The presence of hydrogen on the electrodes also influences the workings of a circuit, partly increasing the resistance and partly making the electrode more electropositive. It is, therefore, not surprising that the galvanic actions either do not occur, or remain unnoticed when the electromotive forces are weak. Some allowance should therefore be made for the density of a current; this is measured by dividing its intensity by the sectional area of any particular point of a circuit; and it is of interest to note that the denser it is at the negative electrode, the more ozone instead of oxygen is generated there.

Heat and Galvanic Currents.--Experiments seem to show that when heat travels along a conductor, an electric current is set up which travels in the same direction. It is however difficult to see how a proper circuit can be established in a boiler, for the electricity would find too much resistance if it travelled from the flame to the furnace plate, thence into the water, into the boiler shell, and back through the air into the furnace. If however such currents do exist, then all heating surfaces should corrode whether made of iron or copper. Practical experience does not favour this theory.

Galvanic Action of Black Scale.--As regards the possible electric actions inside a boiler, there are currents passing from the exposed iron through the water to patches of slag and back, tending to dissolve the iron and to make it as electronegative as the slag, when all action ought to cease.

« AnteriorContinuar »