The Corliss or semi-rotating valve had proved its serviceability, its relative ease of operation, and its tightness under long continued wear, by almost half a century of successful use in the steam engine, at the time when the development here under discussion began, therefore it is not surprising that this valve should have been one of the first applied to air cylinders. In its earlier form it was applied in almost the identical way that it was in the steam engine, except that the inlet valve in the air cylinder corresponds to the exhaust valve in the steam cylinder, as does the discharge valve of the air cylinder with the inlet valve of the steam cylinder.

A glance at the indicator diagrams, Figs. 73 and 75 showing steam cards, Figs. 74 and 76 showing air cards, will show that these are of the same general type and shape, but one is traced in one direction and the other in the opposite. One represents positive work and the other represents negative work, or energy absorbed. So the base line of one represents spent fluid discharged, and of the other incoming fluid to be compressed.

The inlet line of the air cylinder changes but little with changing conditions, that is to say, the times at which the inlet valve opens and closes are only slightly altered by changes in speed or pressure.

Theoretically the proper time of opening is changed slightly because the greater quantity of air in the clearance space at higher pressures expands to a greater volume at atmospheric pressure, and in order to recover all the energy due to the complete re-expansion of this air the inlet valve should not open until re-expansion brings the pressure of this clearance air down to that of atmosphere, which, of course, means that the inlet valve should open somewhat later with high pressures than with low. But no practical or even appreciable loss is experienced under any ordinary condition, by having this point of opening fixed at the point to suit normal conditions.

Fig. 73.

Fig. 74.

Fig. 75.

Fig. 76.

Figs. 73-76. Steam and air cards.

For these reasons the inlet valve is positively connected to the eccentric and opens and closes at predetermined points in the stroke. The discharge valve on the other hand ought obviously to open at a different point with four-pound pressure from what it would for twenty, and it was long thought that it was necessary to put the opening of this valve under the control of some type of trip mechanism, akin in principle to the Corliss dash-pot, which was released by a small auxiliary cylinder in direct communication with the end of the main cylinder, so that as the pressure in the latter rose, the auxiliary piston tripped the releasing gear and the dash-pot pulled the valve open.

It was left for Mr. B. V. Nordberg, of the Nordberg Manufacturing Company, to point out that the loss by having the discharge open a trifle too soon was almost insignificant, in fact practically inappreciable even with very considerable changes in the discharge pressure, and he therefore builds blowing engines, in which both the inlet and the discharge valves are rigidly connected to eccentrics, timed to open and close them at the proper point for normal conditions. A clause in the contract for the first of these engines provided that if it should be found necessary, an automatic gear would be provided at small additional expense which would shift the point of opening of the discharge valves as the pressure rose, but the results from the engine in practice were almost absolutely perfect, and no conditions ever arose to cause regret at the omission of this automatic device.

Corliss valves have several great advantages. By a suitable design of valve gear they can be made to perform most of their travel at a time when the pressure on their opposite sides is approximately the same, so that they can be moved with a minimum of power and of wear, and owing to the "dead travel" obtainable with a proper design of wrist plate are almost stationary when the unbalanced pressure upon them is considerable. They are held against their seats by pressure, and follow them down as they wear, with the result that their tightness in service under good conditions is well maintained. They also permit large port areas with minimum clearance.

On the other hand they require a considerable quantity of oil for proper lubrication, and one large set of engines with this type of valves so designed as to require an excessive amount of oil to keep them in operation was probably more responsible than any other one condition for their failure to develop into a leading type of valve. In spite of this fact at least one engine with valves of this type is known to the writer which has been in virtually continual operation for twenty-five years and is in first-class condition now.