took several indicator cards directly from the valve itself. A small brass rod (") came through a stuffing-box directly over the center of the valve, and over this was placed an old-fashioned MacNaught indicator with a light spring. Just how much this stuffing-box and spring affected the correctness of the diagram it is impossible to say, but care was taken that at the instant of taking the cards the stuffing-box would be very free. These cards are reproduced at full size in Figs. 121, 122, and 123, and the dotted lines are added to represent the line of velocities of the plunger at all points of the stroke. Fig. 124 is an enlargement of a card taken at 8.75 revolutions per minute in order to show more perfectly the relation of the two lines. It is not possible to know the exact diameters at which the discharge may be considered to take place, but I assumed it for the lower seat at 12 inches, and the upper at 9 inches. 11 rev. = .1760 sq. ft. = .2585 1.576 66 4 × 2 × 11 × 1.57 ÷ 60 = 2.30 ft. per sec. Weight of valve per square inch of unbalanced area, 1.00 Ratio of weight of valve to pressure due to flow through the valve, about 1 to 3. The diagrams, as well as experience, showed : First, that the width of a valve seat could safely be brought to a very narrow surface, probably much less than I made it (''); for the pressure in this case was but 377 lbs. per square inch of surface. Secondly, that the lift of a valve is exactly proportioned to the velocity of the plunger, if it is not too light so as to be brought to its stop before the maximum velocity of plunger is attained. The deviation from this theoretical curve, as shown in the cards, is attributable to the friction of the stem running to the indicator, and possibly somewhat to seat area, small as it is. Thirdly, that in the form of valve shown, the theory that the velocity of the water through the valve is that due to the head corresponding to the weight of the valve per square inch of unbalanced area, did not prove to be very near the truth. I can conceive of but one reason for this great variation from the theory assumed, and that is, the effect of the horizontal issuing stream diminished the vertical pressure. I think it is not improbable that there is a mathematical demonstration for the resultant vertical force due to an issuing horizontal stream from a curved aperture, and express correctly the relation of lift to weight, but I have not attempted to thus solve that problem. I do not think that the friction of stem or force of spring is sufficient to account for the deviation. The valves were symmetrical and round in form, and were afterwards turned down and reduced in weight to 35 lbs. in water, or .66 lbs. per square inch of inside unbalanced area, but I regret that I took no further diagrams. The narrow seats, and the quiet action, and the synchronous motion with the plunger, were the more important features in my mind at that time, and the question of weight of valve was left to experiment after all. OTHER DOUBLE-BEAT PUMP VALVES. The valves for the Cornish Pumping Engine at Providence, R. I., were made of substantially the same design as the one just described for the High Service Engine. The dimensions and form are shown in Fig. 125. Its seats were chamfered in the same manner to about one-eighth of inch in width, and the valves always worked well, although nothing is known of the pressure required to operate them. The valve weighed in water, per square inch of unbalanced inside area, 1.28 lbs.; outside area, 1.11 lbs.; and seat area about 12 per cent. of inside area. Water pressure 80 lbs. Pressure upon seats, 680 lbs. per square inch. ST. LOUIS HIGH SERVICE PUMP VALVE. I do not know who designed the valves for this engine, but I am in possession of a drawing of one, which is shown in Fig. 126. From the data given I find that its weight in water per square inch of inside unbalanced area is 1.86 lbs. ; and for outside unbalanced area is 1.12 lbs. On the drawing is this indorsement: "These valves are working under a pressure of 90 lbs. per square inch, seat with very little noise, and give perfect satisfaction at the St. Louis Water Works, November 27th, 1873." Seat area is 67 per cent. of inside area, and pressure per square inch about 250 lbs. MILWAUKEE WATER WORKS PUMP VALVE. I am indebted to Mr. Edwin Reynolds, General Superintendent of the E. P. Allis & Co. works, and a member of this Society, for a full description and drawing of this valve. It is reproduced to scale in Fig. 127. Its weight in water per square inch of inside unbalanced area is 40 lbs.; outside unbalanced area is .21 lbs. |