Opinion
No. 8729.
January 13, 1942.
Appeal from the District Court of the United States for the Eastern District of Michigan, Southern Division; Arthur J. Tuttle, Judge.
Patent infringement suit by Otto C. Berry and another against the Bohn Aluminum Brass Corporation and another. From a decree of the District Court in favor of plaintiffs, 29 F. Supp. 516, defendants appeal.
Decree reversed and cause remanded, with instructions to dismiss the bill.
Arthur C. Denison, of Cleveland, Ohio, and George F. Scull, of New York City (Arthur C. Denison, of Cleveland, Ohio, and George F. Scull and Frank J. Kent, both of New York City, and Karl B. Lutz, of Detroit, Mich., on the brief), for appellants.
John H. Bruninga, of St. Louis, Mo. (John H. Bruninga, of St. Louis, Mo., and Alexander F. Baillio, of Detroit, Mich., on the brief), for appellees.
Before SIMONS, ALLEN, and HAMILTON, Circuit Judges.
This appeal involves Berry Patent No. 1,877,490, application filed March 20, 1922, Claims 5, 15, 18 and 35 of which the District Court held valid and infringed, ordering an accounting and granting the usual injunction. Appellee Berry is the author of the device disclosed by the patent and owner of the legal title; appellee General Motors Corporation is the exclusive licensee under the patent. The accused pistons are manufactured by appellant corporation (hereinafter referred to as Bohn) in accordance with patents issued to and designs made by appellant Nelson. The District Court found that three types of the Bohn pistons, the Packard narrow-strut, the Dodge wide-strut, and the Graham Autothermic, infringe the patent. While the Packard narrow-strut piston is no longer used in automobile production, for a number of years it was used by some of the largest manufacturers of motor cars, about twenty million being made by Bohn itself. The Dodge wide-strut type replaced the Packard narrow-strut, and about thirteen million of the wide-strut type, and some two million of the Autothermic have been manufactured. No commercial piston has ever been made under the patent in suit.
Claim 5.
"A piston, comprising a head-part of non-ferrous metal, wrist-pin bosses rigidly connected to said head-part, skirt segments, and members made of a metal having a thermal co-efficient of expansion smaller than that of the metal in said head-part, the extreme ends of said members being imbedded in said skirt-segments, and intermediate portions thereof being imbedded in the wrist-pin bosses, said members connecting said skirt-segments together and to said wrist-pin bosses."
Claim 15.
"In a piston, a head, a separate sectional skirt of aluminum alloy, and steel struts extending transversely of the piston pin bearings uniting the sections to each other and to the head, said struts having their opposite ends buried in the respective skirt sections, substantially as set forth."
Claim 18.
"In a piston, a sectional skirt, a head separate from the sections of the skirt, said head and skirt consisting of expansible material, and struts of relatively in-expansible material uniting said sections of the skirt to the head and to each other, said struts extending transversely of the piston pin bearings and having their edges adjacent said piston pin bearings buried in the material of said bearings, substantially as set forth."
Claim 35.
"A piston having a head, piers depending from the head, piston pin bearings formed in the piers, cylinder-bearing portions on opposite sides of the piston, a pair of chordal struts, each of said struts extending between the cylinder-bearing portions, the intermediate portion of each strut being surrounded by the material of a pier, there being an integral connection between each of the cylinder-bearing portions and each pier."
Both Berry and Nelson for many years have been concerned with an attempt to solve the problem of the thermal expansion of the aluminum piston in the cast-iron cylinder of the automobile engine. Each claimed to have been the first to make the structure substantially as shown in the patent in suit. In various proceedings before the Patent Office, including interference proceedings between Nelson, Berry and Jardine, priority was adjudged in favor of Berry by all of the Patent Office tribunals, by the Court of Customs and Patent Appeals, and by the District Court in a decree affirmed by this court. Cleveland Trust Co. v. Berry, 6 Cir., 99 F.2d 517. Whatever persuasive force these holdings have, they are not controlling when the validity of the claims is involved in an infringement case. Schriber-Schroth v. Cleveland Trust Co., 305 U.S. 47, 59, 59 S. Ct. 8, 83 L.Ed. 34.
The Berry piston is a so-called "slipper piston," composed of the usual head and a skirt, divided preferably into two opposite skirt segments. The head discloses a ring belt for piston rings, and the skirt segments are completely separated from the head by a gap. Piers running down from the head between the skirts connect the head with the wrist-pin bosses, which lie between the skirt segments. All parts of the piston are aluminum with the exception of two wide corrugated iron struts which have their ends embedded near the edges of the skirt segments and their middle portion attached to the wrist-pin bosses. The struts are carried to the full extended length of the skirt. The steel or iron struts have a melting point higher than that of aluminum, and therefore the specifications suggest that the struts be cast in place (a process old in the art) in order to secure a stout and integral union with the bosses and the skirt. This is achieved by positioning the struts in a permanent mold adapted for casting the piston and pouring molten metal for the aluminum parts of the piston around the struts.
Appellants claim that the accused devices do not infringe. As to the Autothermic piston, they urge that since the coefficient of expansion of the metal used in the strut is greater than that of cast iron, it is not covered by Berry. The Autothermic strut is made of mild steel, which has a coefficient of expansion of 6.3, while that of aluminum is 11.1 (these figures being agreed upon by the parties for purposes of comparison). It is true that Berry's specification calls for "some strong metal having a coefficient of expansion not greater than that of the metal of the cylinder" and preferably "smaller" than that of the cylinder metal. Since the cylinder is of cast iron, which has a coefficient of expansion of 5.6, the Autothermic strut is more expansible than the metal of the cylinder. However, in Claim 5 Berry specifies that the struts shall have a coefficient of expansion "smaller than that of the metal" in the head. Since the metal in the head is aluminum, this clearly means that the struts shall have a coefficient of expansion smaller than aluminum. Upon this feature Berry reads on the Autothermic piston.
Appellants also claim that the accused pistons are not within the Berry patent because the steel in the three pistons is under compression, while this is not the case with Berry. This, we think, is a distinction without a difference. The record shows that the accused struts perform the equivalent function of the struts of Berry in restraining the expansion of the aluminum skirt and making possible a very small clearance between piston and cylinder to allow for expansion under heat. The fact that the skirt of the Autothermic piston is continuous instead of in segments, as specified by Claim 5 of Berry, is also an immaterial difference. Cf. Cleveland Trust Co. v. Berry, supra, 99 F.2d at page 522. We conclude that the District Court correctly found that the accused pistons infringe if the Berry patent is valid.
Claim 35 is clearly invalid. While it describes the structure of the piston, at no point does it specify that the chordal struts which connect the skirt segments shall have a coefficient of expansion less than that of aluminum. This feature is either the gist of the invention, or at least an essential part of the combination claimed to be patentable, and its omission is fatal. As written, the claim applies to an all-aluminum piston as well as to the device described in the specifications and is too broad to be valid.
The principal question presented, then, is the validity of Claims 5, 15 and 18 of the patent. These claims, which present no essential difference, are printed in the margin. The claimed object of the invention was "to produce an internal-combustion-engine piston which has all the advantages of non-ferrous pistons, but which is free from the disadvantages usually incident to non-ferrous pistons, such as excessive expansion upon heating." The non-ferrous alloys are lighter, have a lower coefficient of friction, a higher thermal conductivity, and are usually softer than cast iron. But they have a higher coefficient of expansion than iron, aluminum expanding more than twice as rapidly as cast iron. An aluminum piston therefore expands in an iron cylinder to such an extent that if a proper allowance is made for expansion of the piston so that the cylinder will not score, the piston is too loose when the engine is cold, causing what is known as "slap." Both the scoring of the cylinder and the slapping of the piston in the cylinder are harmful, and the problem of controlling the expansibility of aluminum has challenged the engineering skill of the automotive industry for a long period. The subject has been before this court in repeated litigation. Cleveland Trust Co. v. Schriber-Schroth Co., 6 Cir., 92 F.2d 330; Id., 6 Cir., 108 F.2d 109; Cleveland Trust Co. v. Berry, supra; Aluminum Co. of America v. Thompson Products, 6 Cir., 122 F.2d 796.
Berry claimed to solve the problem by employing corrugated sheet metal members or struts of a metal having a coefficient of expansion preferably smaller than that of the cylinder metal to compensate for the fact that the piston heats up more rapidly than the cylinder. He claimed to exert this control at the top of the skirt of the piston which, when the engine is running, is higher in temperature than the bottom of the skirt because it is nearer the point of the explosions in the cylinder. Appellees' expert, Louis M. Stellman, on direct examination testified that the use of struts of steel was the principle of the Berry piston. Differentiating it from the Maxwell piston and the Long piston, which are all-aluminum pistons, he stated that in Maxwell and Long there was expansion along the aluminum struts in the direction of the struts, while in Berry the expansion was at right angles thereto. Stellman said that Berry showed very little expansion along the strut because it was made of steel, and that this held the thrust surfaces in instead of allowing them to expand.
The District Court found that this principle was understood in the art and that the idea of inserting a steel strut or support to restrain the expansibility of the aluminum was not new, but he found that the prior Nelson single, two and four-iron strut pistons in which the strut was of iron and the rest of the piston of aluminum were not workable. Bearing in mind the 400-pound side thrust, which results as the piston reciprocates in the cylinder, the court considered that if the thrust surface is held by one or more struts pressing directly against the thrust surface, both the thrust surface and the support for the thrust surface are "undesirable." The court declared that the skirt should be cam ground in order to have it work properly. If properly cam ground and held at both ends by struts, the court found that the Berry surface was a good thrust surface. The essential to securing this result the court stated was to have the Berry skirt oval or cam ground, and assuming the presence of this essential step, he declared the patent valid. But this was not disclosed by Berry. As Stellman said, Berry "didn't say anything in his patent about that." The drawing of the Berry patent discloses not an oval, but a round thrust surface with its support at the edges, the two struts passing through the bosses and striking the thrust surfaces near the extremities of the skirts. The court found that this essential feature was disclosed in the patent specification by the statement that "The rough piston thus produced is then finished in the usual way of pistons, to provide proper clearance for both the skirt-segments and the head-part." But this sentence is otherwise explained by the sentence of the specifications immediately following: "As already stated, this clearance is relatively large for the head-part, to allow for the differential expansion of it and the cylinder, but is relatively small for the skirt section, there being just sufficient clearance to permit proper lubrication." Concededly the clearance for the head must be larger than that for the skirt because the head receives the immediate heat engendered by the explosion, and securing such a clearance is in fact the usual practice. The phrase "finished in the usual way of pistons" as defined in the specifications, means finished with a clearance larger for the head than for the skirt, and in no way discloses nor even suggests cam grinding as an essential feature of the structure.
In fact this record does not show that it was usual that pistons be cam ground. Berry himself throughout the long involved interference proceedings never suggested cam grinding. The only patent introduced which shows an oval grinding (Mason and Sturt, No. 1,275,064) discloses no struts at all. Evidently oval grinding is not commonly employed in pistons, as found by the District Court, for the purpose of "getting a relief over the ends of the struts." When relief was desired, the actual practice was to cut away the thrust surface to a uniform depth over the ends of the struts. This is the well-known "keystone" relief. Cf. Cleveland Trust Co. v. Schriber-Schroth, 6 Cir., 92 F.2d 330, 337. The District Court's finding that an aluminum piston in which the thrust surface is held by each strut pressing directly against the thrust surface, is unworkable is disproved by this record, for the Packard narrow-strut piston, which is not ground oval and in which the edges of the thrust faces bear directly against the cylinder, has been used in twenty million engines by highly important automobile producers.
We conclude that the District Court read into the claims in issue as an essential element the requirement that the thrust surface should be "cam ground." This was not revealed in the drawings, the specifications, or the claims, and therefore the patent held to be valid was neither claimed nor granted. Altoona Publix Theatres v. American Tri-Ergon Corp., 294 U.S. 477, 487, 55 S.Ct. 455, 79 L.Ed. 1005; Schriber-Schroth v. Cleveland Trust Co., supra; Permutit Co. v. Graver Corp., 284 U.S. 52, 52 S.Ct. 53, 76 L.Ed. 163.
It remains to consider the claimed invention revealed in the claims in issue, the principal feature of which, as reiterated by Berry's expert, was the insertion of the struts of steel in the aluminum piston. Speaking of the difference between the Berry piston and the piston of Ricardo, (which was in evidence in Cleveland Trust Co. v. Schriber-Schroth, 6 Cir., 92 F.2d 330), and other strut pistons, Stellman said that "these struts are made of steel instead of aluminum, and they do not have the expansion, as much expansion as the aluminum struts, and therefore the skirts can be fitted closer and do not come in contact with the cylinder, that is at the same temperature." Differentiating from the Maxwell piston, he said, "The struts in the Maxwell piston are of aluminum and in the Berry piston are of steel. So that we get the expansion of aluminum as compared with steel in the two pistons. Fundamentally that is all there is to it, I think." When the court asked, "The only difference is as to the material used?" Stellman answered, "That is right," and when the court asked "The only difference in operation is when heated, involving that principle of heat," the answer was "That is right. . . . It means that you can run the Berry piston with less clearance between the cylinder and the piston than you can the Maxwell."
But this feature was not new and had been amply revealed in the prior art. It was so found by the District Court. Knight, 1,418,719, discloses an aluminum piston operating in a cast-iron cylinder with a ring in the skirt made of cast-iron or of the same material as the cylinder so as to control the expansion of the piston. This specific feature is set forth in the claims. Napier, 1,525,316, also describes a control member formed of a metal with a "coefficient of expansion different" from that of the skirt. Hartog, 1,794,767, seeks to control the excess expansion of the piston by use of a member having a different ratio of expansion of "less capacity" than that of the material out of which the piston is made, and claims this as one of the essential features of his device. Nelson's patent, 1,454,761, for a single strut piston, application filed October 27, 1920, presents a piston with a strut extending transversely of and above the piston pin, and suggests that it may be of different material cast in place. Nelson also described this conception in an article published in the Journal of Automotive Engineers in February, 1921, with reference to the constant-clearance aluminum piston, in which he says:
"* * * it has become a habit to think that aluminum pistons must expand. Why not design an aluminum piston that cannot expand as far as cylinder clearance is concerned? This viewpoint changed the complexion of the whole problem and led to the development of aluminum automobile piston design that gives results thought to be in advance of the combined merits of the aviation piston and the cast-iron piston.
"* * * In this design the adjustable steel strut controls the cylinder clearance in the direction which prevents piston slap."
Berry admits that he heard Nelson read this paper at a meeting of the Society of Automotive Engineers in January, 1921. We conclude that the substitution of the metal, under this record, does not constitute invention.
The District Court declared that it was necessary to move the iron struts "into a new place" in order to perform the desired service. Assuming this to be true, the change of position disclosed in Berry does not involve invention. In view of Berry's separation of the skirt segments from the head, the only possible place to put these two struts where connection could be made with the piers was at the ends of the skirts. Otherwise the skirts would not have been fastened to the head. Also, as appellees' expert said, it was logical for them to be placed there to avoid interference with the connecting rod. Finally, and we think conclusively upon the question of invalidity, appellees' contention is shown to have no merit by the undisputed fact that Berry placed the struts in exactly the same position as the supporting webs of Ricardo, whose slipper piston he was imitating. Berry conceded that the slipper type Ricardo piston was much in favor at the time he applied for his patent, and said "we followed, or sort of copied him in that respect." So far as the structure of the skirt is concerned, Ricardo is very similar to Berry, the skirt segments being connected by webs of aluminum which are cast integral with the wristpin bosses. The only substantial distinction between Berry and Ricardo lies in a substitution of the metals, which was old.
The feature of embedding the struts in the bosses is also claimed to constitute patentable invention. The Patent Office held Claim 15 valid over Nelson's article on the ground that struts "uniting the sections to each other and to the head" had not been disclosed therein. In Hartog, however, there was an aluminum piston with a steel control member connected to the bosses and to the skirts. This patent is of particular significance because it was not considered in the Patent Office.
Much testimony was taken here in an attempt to establish a new mode of operation by reason of using chordal struts and imbedding them in the bosses. It was claimed that the factor of safety was materially increased in Berry's device. But if we were to assume that the element of imbedding the struts in the bosses was novel, it would not help appellee's case. Stellman was of the opinion that the added strength came from using two struts instead of one, and that embedment of the struts in the bosses did not have "so much to do with it." If, as he stated, the greater strength was due to "the mere fact that you have got steel struts separated from each other," there is no new mode of operation in Berry for that had been shown by Nelson, and as Stellman admitted, it was the "logical thing to do" to escape interference with the connecting rod.
Although the Ricardo piston, which shows aluminum webs or struts joining the skirt sections to each other and to the head, was before the Patent Office, it held that this patent did not negative invention because it was not "obvious" how the Ricardo webs could be made out of iron "in a composite structure of different metals." The present record establishes that the process of inserting a metal with a high melting point in casting a metal having a lower melting point, technically referred to as casting in place, was well known to the art. Indeed such a construction was mentioned by Nelson in his single strut piston patent as an alternative method of inserting the single strut.
It follows that the District Court erred in finding the claims in suit to be valid.
The decree is reversed and the case is remanded with instructions to dismiss the bill.