ee  eOs8“aOee“qeq73o3oewe0SS SS ><a 
THE CORROSION OF CONDENSER TUBES. 
— — 
running sea water have shown that a period of fifteen to twenty years 
may elapse—even presuming that the rate of corrosion is uniform and 
does not decrease with time—before a tube is reduced 50 per cent. in 
thickness from this cause. Tubes almost always fail in practice by 
loeal action of some kind long before ordinary complete corrosion has 
seriously reduced the thickness of the tube. Rapid general thinning, 
however, is essentially a fresh-water phenomenon, and is usually asso- 
ciated with the presence of free acid in the water supply. The results 
of tests on ten samples of tubes of widely different composition in 
hydrochlorie acid of concentration 3 parts in 100,000 at ordinary tem- 
perature show that in six weeks all tubes had lost from 2 to 4 per 
cent. in thickness. That such a very dilute acid should reduce the 
thickness of a tube so much in such a short time at ordinary tempera- 
ture is ample evidence of the serious effect of acid in the water supply 
upon the life of condenser tubes. Proper neutralization, preferably at 
the souree of contamination, is an effective remedy for this trouble, , 
but considerable difficulty may be experienced in detecting the acid, 
especially if, as is generally the case, it enters the water supply only 
intermittently. Regular and frequent tests of the water must be made 
whenever this type of trouble occurs. 
Type Il., Deposit Attack.—The principal cause of pitting, which 
is the most frequent source of trouble in condensers, is ascribed to what 
is termed “deposit attack.’ In the presence of sodium chloride solu- 
tions, the cuprous oxide formed on a brass surface gradually changes 
to cuprous chloride. The latter is usually swept out of a condenser 
tube by the circulating water, but under various conditions may adhere 
at different parts of the tube surface. When such adherence has 
occurred, conditions now allow of the further gradual change of the 
insoluble cuprous chloride under the influence of oxygen to soluble 
cupric chloride and euprous oxide. The action of cupric chloride 
solution on brass is very rapid, as may be gathered from the fact that 
a piece of brass tube, 2 inches long, placed in a strong cupric chloride 
solution at ordinary temperature, was completely disintegrated and 
partially replaced by a pseudomorph in copper in two days. The 
action involves the oxidation of the copper and reduction of cupric 
chloride to cuprous chloride. | Redeposition of copper from solution 
by the zine also occurs. In the presence of air cuprous chloride will 
again be converted to cupric chloride and the attack on the brass con- 
tinued. Thus the action is both recurrent and local. Foreign bodies, 
and particularly colloidal bodies, lying in the tube have an injurious 
effect by serving as loci for the adhesion of cuprous chloride and by 
preventing the diffusion of cuprie chloride. Observations on the in- 
cidence and distribution of pits in condenser tubes are shown to agree 
with the results which would be expected to follow from the above ex- 
planation of the mechanism of pitting. Attention is drawn to the 
importance of keeping tubes clean and as free as possible from foreign 
bodies as a means of preventing deposit attack. ‘seg ; 
Type IIL., Layer Dezincification—An account is given, preliminary 
to the consideration of this type of corrosion, of the mechanism of so- 
called dezincification. The conelusion is reached that true parting 
of zinc and copper in a 70 : 30 brass does not occur, but that the 
357 
