Opinion
No. 7831.
December 11, 1939.
Appeal from the District Court of the United States for the Northern District of Ohio, Eastern Division; Paul Jones, Judge.
Patent infringement suit by Harry E. La Bour and another against the Gorman-Rupp Company. From a decree for defendant, plaintiffs appeal.
Affirmed.
Frank Parker Davis and John A. Dienner, both of Chicago, Ill. (John H. Leonard, of Cleveland, Ohio, and Edward C. Grelle and John A. Dienner, both of Chicago, Ill., on the brief), for appellants.
Arthur C. Denison, of Cleveland, Ohio (Baker, Hostetler, Sidlow Patterson and A.C. Denison, all of Cleveland, Ohio, Benjamin B. Dowell, of Washington, D.C., and Osgood H. Dowell, of Chicago, Ill., on the brief), for appellee.
Before HICKS, SIMONS, and ARANT, Circuit Judges.
Harry E. LaBour, as patentee, and The LaBour Company, Inc., as exclusive licensee, sued Gorman-Rupp Company, a pump manufacturer, for infringement of thirteen claims of Patent No. 1,578,236, March 23, 1926, to LaBour for a "Centrifugal Pump." The chief defenses were, lack of invention and noninfringement. Claims 21, 24, 32 and 41 only, each for a method, are involved.
The District Court found no infringement.
The claims provide a method for the automatic priming of centrifugal pumps. The priming operation, that is, supplying water or other liquid to a pump to exhaust the air in the pump housing and piping, preliminary to actual pumping, is more or less familiar.
Self-priming centrifugal pumps, which operated through re-circulation of water caught in a trap in the intake pipe and controlled by hand, or by automatic valves, were old. But the hand control system required personal attention and both systems were susceptible to clogging and fouling in the valves where used in operations requiring the passage of sand, pebbles and trash through the pump.
The LaBour method re-circulated the priming water automatically without the use of valves and LaBour claimed that the re-circulation stopped as soon as the priming was completed. Earlier priming operations re-circulated the water between the discharge pipe and the intake. LaBour shortened this return by running it from an outlet in a dome-shaped enlargement in the discharge pipe called the "separator," located just above the impeller through an auxiliary inlet, "entering the pump near the outer portion of the impeller * * * at a point adjacent the outlet but posterior thereto with regard to the direction of rotation of the * * * impeller."
Touching this inlet, he states in his specification, "I have an auxiliary inlet port which operates during the priming period as a means for permitting the liquid to reenter the pump to maintain the same primed, but which becomes substantially ineffective as an inlet port and in fact may serve as an outlet port when the pump is operating upon liquid in normal operation. It is thereby possible to secure a maximum priming action when necessary, but a minimum loss when the pump is operating under normal conditions." Again, that the "advanced edge of the auxiliary inlet port 27 is substantially radial, while the trailing edge is rounded. The result is that a slug of liquid finds difficulty in moving out of the pocket" [of the impeller] "into this auxiliary inlet port when the pump is full of liquid."
It seems clear that when the patent was issued LaBour thought that re-circulation through the inlet port ceased when priming was complete; but whether there was a reverse flow was problematical. However, at the trial he was more certain of the point. He claimed, and appellants now claim, that the provisions for peripheral reentry of the fluid into the impeller make use of hydraulic balance; that is, as the air tends to eliminate, the centrifugal pressure from the whirling liquid becomes greater, reversing the flow in the reentry passage and eliminating the need for valves, and preventing clogging by the movement of current both ways.
LaBour's impeller was "an open pump runner, having an open spider with a ring upon its outer periphery, this ring having transverse blades or impellers which project radially and laterally from the side of the ring. * * *" The impeller moved counter-clockwise and was operated in a casing concentric with it. LaBour thought the concentric shape of the housing was important. He said in the specification: "It is desirable that the blades or vanes extend closely adjacent to the bottom of the channel 42 in order to secure maximum effect. It is not desirable so far as I have found to have a radial flow of liquid through the pockets, as is customary in the ordinary volute pump, except during that period when the pockets register with the discharge opening."
LaBour's theory, as stated in appellants' brief, is, that the rapidly rotating tips of the impeller strike the water and break it into a spray or mixture with the air in the intake pipe; that "this spray or mixture of air and water is swirled around the inside of the pump to develop centrifugal force. Then as it reaches the throat" [outlet] "it is driven out in the separator. The spray or mixture thus carries out air along with the water. Here it slows down and the air rises to the surface. The water without air settles to the bottom and some of it returns through the passage-way * * * into contact with the blades of the impeller. Thus, by continuous action, air is carried out. The carrying out of air creates a suction in the suction pipe, and sucks water up higher and higher as the suction increases. As the suction increases, the return of priming water into the periphery is increased, and a denser or heavier spray or mixture of air and water is formed. This is necessary to produce a discharge pressure required to drive the mixture out against this opposition of the suction produced in the pump casing and suction pipe."
Claim 32 is as follows:
"32. The method of operating a pump of the class described having a casing and an impeller to cause it to pick up its own liquid load, which comprises rotating the impeller, mixing liquid with the air in the casing of the pump by rotation of the impeller, moving the mixture along the periphery of the casing, throwing out a stream of the mixture by centrifugal force at a suitable discharge point, separating the liquid and gas under discharge pressure, returning a part of the liquid into the periphery of the casing posterior to the point of discharge and stopping the reentry of said liquid into the casing by the increased centrifugal force developed by liquid free of gas." (Italics ours.)
The relevant portions of the other claims follow:
"21. * * * and guiding the liquid back peripherally into the impeller chamber. * * *
"24. * * * and returning the separated liquid peripherally into the impeller chamber. * * *
"41. * * * and returning the liquid to the impeller to be mixed with gas in the pump and discharged out of the discharge orifice until the suction pipe is evacuated. * * *" (All italics ours.)
Appellee's pump, like LaBour's, had a trap and separating chamber but the chamber was flatter and more bowl-shaped than LaBour's. Appellee's impeller was three bladed, with the blades greatly curved backward and overlapping each other by about half their length. The movement was clockwise. The casing was volute, inducing some radial flow of the liquid before it reached the outlet port. This port was directly over the axis of the propeller, so that the discharge, as in LaBour, was upwardly. The edges of this outlet were in the curve of the volute on the advance side and just outside the path of the impeller on the posterior side. The floor of the separator was divided into two parts by the outlet opening, so that a shelf was formed on either side thereof, the one on the posterior side being lower by the width of the volute than that on the advance side.
Appellee denies that there is any re-circulation of water in the operation of its pump except for an inappreciable and unavoidable leakage which takes place between the impeller and its casing and which is common to all centrifugal pumps discharging upwardly. Its theory, as stated in its brief, is that its pump "utilizes the priming water in a manner to provide a liquid seal of low head sustained on the rotating propeller directly over its axis and substantially or nearly around it, enabling the impeller to discharge the air from substantially all around its periphery into the surrounding water * * * the pump impeller has every vane overlapped by a prolonged trailing extension of the next, providing deep pockets under and covered by the vane extensions. Such an impeller rotating at high speed exerts outward pressure against the surrounding water while discharging air under centrifugal pressure into the surrounding water."
The explanation of the operation of appellee's pump is set forth in the specification of a patent issued to Rupp, No. 2,002,454, May 21, 1935, where it is said that "the cooperative chambers and passages are so arranged in relation to each other that the customary re-circulation of liquid, — from the discharge to the suction side, — or an auxiliary evacuating means, are eliminated."
The peripheral reentry of the liquid into the impeller chamber is the important step of appellants' method. Three of the claims expressly call for peripheral reentry and the fourth by implication. Our question is, whether appellee infringed this step. The District Court held that it did not. Our conclusion is that appellants have failed to carry the burden of establishing, by the weight of the evidence, that it did.
Appellants' evidence consisted largely of expert testimony chiefly with reference to moving pictures made ex parte; of the physical operation of appellee's pump and LaBour's explanation thereof and the file wrapper of the Rupp patent. Some of the demonstrations were repeated before us. The fact that the moving pictures were made without notice to appellee is to be taken into consideration in determining the weight to be given thereto. But, apart from this, the pictures themselves are not convincing. There was one scene showing pieces of rubber, of greater specific gravity than water, floating on the surface of the water in the separator. Some of these, while moving, disappeared, as appellants' expert McNulty, stated, "toward the impeller." Admittedly, there was turbulence but no positive showing, by the use of the rubber bands, of any downward current through the outlet. In a demonstration before the District Court, LaBour inserted into the separator a rod with strings on it. He stated that the impeller had pulled in and cut off pieces of string, but the details of this experiment are very meager and did not impress the District Court. The court's observation was made through windows specially arranged in the pump mechanism. The experiment was made, of course, for the purpose of showing that water was going down into the impeller. LaBour frankly admitted that this was difficult to see. This experiment was not repeated before us.
LaBour admitted that the water did not come back into the impeller in strictly confined paths. He stated: "If I may use a simile, it is more like a traffic intersection, with no stop lights and no policemen." But the simile fails because the upward stream of his "traffic" was being forced at terrific speed. The impeller operated at from 1,750 to 1,800 revolutions per minute and we cannot see how there could be any cross movement against that stream. Appellants' showing that the pump also primed at lower speeds is not evidence, without more, of their theory of operation.
In fact, the evidence that the water in appellee's pump re-circulated at all was unconvincing and the results of appellants' demonstrations were indeterminate. It was not shown by the preponderance of the evidence that in appellee's pump a part of the water returned into the periphery of the casing posterior to the point of discharge as in Claim 32. It is difficult to see how the liquid could return to appellee's impeller posterior to the point of discharge, for, unlike appellants' pump, appellee had one outlet only. There was certainly no " guiding the liquid back peripherally into the impeller chamber" as in Claim 21. (Italics ours.) Appellants claim no more for the reentry of the liquid into the impeller than that it occurred in a more or less helter-skelter manner.
It is equally uncertain that any liquid or water returned in any orderly manner to appellee's impeller to be there mixed with gas, as in Claim 41.
Finally, infringement is a question of fact and we must rightly give much weight to the opinion of the trial judge who heard the witnesses and saw their manner and demeanor upon the stand.
Assuming the validity of the claims, the decree is affirmed upon the ground that infringement thereof has not been shown by the quantum of proof the law requires.