Some important information in regard to ' the concussion of pump valves was communicated by W. G. Armstrong, in a paper read before the " Assoc. Inst. C. E.," ot which we give below a condensed statement for the be-j, nefit of our readers. In the construction of force pumps, acting under heavy pressures, lor working hydraulic rams and other machines ' great difficulties have been Irequently experienced from the beat or concussion, which has been generally attributed to the fall of the valve upon its seat; a"series of experiments were made with the view ot ascertaining and ' removing this difficulty. On examining the delivering valve, it was evident that while , the entire upper area was acted upon by the downward or closing pressure, only that portion of the under surface which covered the " annular opening was acted upon by the upward or opening pressure. From this observation it was evident that sir.ce the area acted upon from below was much less than that acted upon from above, a momentary excess of pressure must be produced by the plunger in the pump barrel, in order to raise the valve 3 from its seat. The material of the pump barrel would thus be unduly distended, and a sud-I den collapse would take place immediately upon starting the valve. , In the case of the valve which was exhibited, the surface acted upon from below was I only one-sixth ol the surface subjected to j downward pressure, so that the pressure per square incn exerted by the plunger, on start-ing the valve, would be six times the column to be lifted ; all the effects observed agreed with this explanation. For the purpose of further experiment, a valve was constructed , with the annular openings so extended as to diminish the bearing surface, and to lessen the . difference between the areas of the upper and under surfaces ; this construction proved par-i tially successful, and when these areas were nearly equal, a smooth and nearly noiseless ] action was obtained, even though the pump made nearly one hundred strokes per minute. This construction of valves appeared, there-, fore, to be the best, particularly where the pumps were to be worked rapidly. The concussion does not, however, always arise from the cause before stated ; in some cases it is at-, tributable to the valve remaining open until ! after the return of the stroke, and then being suddenly forced down by the weight ot the returning column. The cause of this concussion was probably the excessive rise ot the valve, and the overrunning of the column in the delivery pipe consequent upon the momentum imparted to it by the previous stroke of the pump. Large dimensions were found to be efficacious so as to admit a free passage for the water, and a very limited rise of the valve, and either increasing the diameter of the delivery pipe, or applying the means of keeping up a continuous motion of the column without drawing tail water through the valve.
This article was originally published with the title "Concussion of Pump Valves" in Scientific American 8, 39, 306 (June 1853)