Causes of Heating.

the cheaper. Tin melts at 440 degrees Fahr., and a metal made chiefly from tin should not be overheated. A good rule is that it is sufficiently hot when it will char a pine stick. It should always be covered with a flux when melting to preserve it from oxidation. Charcoal is often used for this purpose.

Heating may be caused by all parts not being lubricated, there being no oil channels to spread the oil; by hard metals made up in the shaft, like pieces of steel, or cast iron, or cinder, or any material that does not wear smoothly and evenly; by the casting not being properly cleaned and sand working out under the lining metal; by the edges of the lining metal not having been trimmed off and the thin edges cracking off; by the work not being in line, or level and the load not distributed evenly; by the journal not being of sufficient size, there being more than 150 pounds pressure to the square inch. In some cases dirt may get in, and in many cases improper lubricants are used. Too tight a belt makes an excess of friction.

Corliss Engines.

This chapter will be to a large extent personal.

For a number of years I had tried to get some one interested in putting on an extra eccentric but was unable to do so, and all Corliss engine builders of the time claimed that it was not necessary and would make a needless combination and expense.

In 1872 I had added to my equipment a Corliss engine, 28 x 60, running at 52 revolutions. To this engine was attached a syphon condenser. At that time indicators were scarce, but I had a Richards. I was unable to get a card that suited me. If the attempt was made to get any compression the exhaust was late and would not show full vacuum before half stroke. I tried compression by giving the eccentric a large advance and by lengthening the exhaust connections. By doing this it was necessary to lengthen the steam connections. This made about three-eighths stroke the latest possible cut-off. As the engine was doing rolling mill work, some of the time it meant full stroke. The slowness of the exhaust also troubled me. It was learned after repeated trials that getting compression at the expense of release meant more coal burned, while the earlier the release, the less coal

Corliss Indicator Diagrams.

was required. It finally settled down to the diagram shown in Fig. 69, as the best that could be done and still have the engine run fairly quiet. I began to talk two eccentrics for Corliss engines, but no one would listen to me, all interested parties claiming the extra one was not needed. I tried to induce those having new engines built to insist on it, but all were easily talked out of it.

In 1883 I was in a position to say that the engine should be changed that way. In conversation one day with the superintendent of the engine works, he was told there was going to be another eccentric. Said the superintendent, "We can build it for you," and it was arranged

Fig. 69. Single eccentric diagram.

that I should send the dimensions and a sketch of what was wanted and the engine builder would make it. It was made in 1884.

For some reason everyone had the idea that the office of a second eccentric was to give freedom to setting the exhaust valves and the principal thing was to get compression. I wanted to get early release and have the vacuum have effect the full length of the stroke; also a longer range of cutting-off.

With a Corliss engine it is evident that the valve must release at or before the full throw of the eccentric, so the steam eccentric was set at right angles to the crank, which would insure a range of cut-off up to half stroke. Both

Wristplates.

eccentrics were set at right angles to the crank, both wrist plates vertical, the steam valves with 1-16 inch lead and the exhaust with 8 inch lead. The exhaust eccentric was then turned to about 30 degrees angular advance of the steam eccentric.

I have always had the idea that one should never depart from the builder's design of an engine if pos

sible; that there should be no special parts, so that repairs could be quickly and cheaply made. The wrist plate originally was like Fig. 70. The new wrist plates were made one-half as thick, with the outline shown by dotted lines, and fitted to the same stud. The new rocker arm was the same as its mate, and all valve connections

More Corliss Cards.

Fig. 71. Diagram from two eccentrics.

were the same. After the new arrangement was started a diagram was like Fig. 71. It will be noticed on this that the cut-off is round. I wrote the builder, sending some cards, and inquired if there was any remedy. The builder suggested that the studs operating the steam valves be set 1 inch nearer the circumference of the wrist plate, which would give the valve more and quicker travel. This was done, and the precaution was taken to work the wrist full throw both ways to see that everything was clear, but when the steam was turned on and the engine was partly up to speed, the dash-pot rod pulled just out of the guide, and the result was a broken wrist plate. As everything was uniform with the old, the old single wrist plate was put back and attached to the steam eccentric set at right angles to the crank, and Fig. 72 was the result.