GENERAL METALS POWDER CO.
v.
S. K. WELLMAN CO. et al.
No. 10086.
Circuit Court of Appeals, Sixth Circuit.
September 23, 1946.*506 F. O. Richey and B. D. Watts, both of Cleveland, Ohio (H. F. McNenny and H. F. Schneider, both of Cleveland, Ohio, on the brief; Richards & Watts, of Cleveland, Ohio, of counsel), for appellant.
George I. Haight, of Chicago, Ill. (Ray Stewart Gehr and C. T. Cox, both of Cleveland, Ohio, and W. E. Lucas, of Chicago, Ill., on the brief), for appellees.
Before HICKS, SIMONS and MILLER, Circuit Judges.
HICKS, Circuit Judge.
Action at law by the General Metals Powder Company, appellant, and assignee-owner, against The S. K. Wellman Company and Samuel K. Wellman, appellees, for damages for infringement of Claims 2, 3, 5, 12, 13, 14, 18 and 19 of Patent No. 2,072,070, for a "Friction Article and Method of Producing Same"; and of Claims 1, 3, 5 and 6 of Patent No. 2,191,460, for an "Article of Manufacture." These patents were issued to Fisher on February 23, 1937 and February 27, 1940, respectively, and are referred to herein as 070 and 460 respectively.
Herein the parties are styled plaintiff and defendants just as they appeared in the court below.
Defendants denied infringement and set up the defenses of invalidity because of prior use and invention; and for lack of invention; and for ambiguity and lack of clarity in the specifications and claims.
The case was tried without a jury. The court pretermitted the question of damages and heard only the issues of validity and infringement. It filed an opinion and findings of fact and conclusions of law. Although remarking in the opinion that before 1934 serious clutch problems existed in the automotive industry and that many manufacturers were striving to devise a clutch lining that would meet adverse use conditions and that Fisher's "friction article" achieved "amazing results" in buses of the Twin Coach Company, the court nevertheless was constrained to find that there was nothing novel in plaintiff's "continuous metal network structure" since the same structure was revealed in defendants' "Exhibit 48" made under the Gilson patent and in the articles made under the Williams and Claus patents. The court found nothing novel in the processes of 070 nor in the application of the composition to the clutch problem since the Hudson Motor Car Company had used clutch brakes made from Genelite material (under the Gilson patent) prior to Fisher. It held the claims of 070 invalid. It also held the claims of 460 invalid because of anticipation by the Short patent No. 1,819,272 and because welding by the heating process and by electrical resistance were old and known prior to Fisher.
Patent 460 was for an article of manufacture such as a clutch plate, brake shoe, bearing or the like, in which a bonded metal element such as indicated above, was "bonded" or "integrally attached" to a steel supporting or backing member.
Plaintiff was a manufacturer of powdered metals and interested in finding outlets therefor. Some experimentation was carried on by its employees, including Fisher, for uses of powdered metal in conjunction with other materials for electric brushes, brake linings, etc. Officers of plaintiff *507 demonstrated to defendants a Ford automobile which had been equipped with brake shoes manufactured by plaintiff and later there was some cooperation and exchange of information between the two companies before they went their separate ways.
Before examining the art and usages claimed by the defendants to antedate the patents, it will be helpful to examine the patents themselves in greater detail.
Patent 070:
This patent related to "improved compositions of materials for use as friction generating media and to processes of making the same * * * friction articles which are particularly well adapted to withstand service conditions encountered in relatively moving bodies in contact with each other, such as in brakes, clutches and similar devices. Such uses require more or less definite but less controllable coefficients of friction, long wearing life, smoothness of operation, and non-abrasive characteristics, together with resistance to deteriorating action or influencing effects of oil. * * *" (Italics ours.)
A feature of the article, emphasized in the trial and in the arguments and briefs, was a "network" of metal in the interstices of which graphite and other non-metallic minerals were dispersed. The specification had this to say about the "network": The friction articles "* * * consist predominantly of a strong, malleable, ductile, wear-resistant metallic network having a small amount of inorganic, non-metallic materials dispersed therein and supported thereby. The term `network' has been selected as an expression which more or less aptly describes the structure of the metallic body of this invention. This structure may be fairly well visualized by a comparison of it with the structure of sponge rubber wherein the rubber is more or less continuous and forms a network with many open spaces. In the friction articles of this invention the metallic parts may be likened to the rubber of the sponge in continuity and in the formation of spaces. Such spaces, howver, contain ingredients which were a part of the starting material but which do not enter into the combination of the metals to make up the network." (Italics ours.)
The specification further stated that copper was the predominant metal in the network material, ranging, according to use, from 65% to 95% by weight, with tin 3% to 15%, and with lead, 5% to 15%. Zinc might be substituted for tin; and iron or cadmium for copper. The inorganic materials, such as graphite, mica, silica, emery, clay, carborundum, etc., were varied to give a different coefficient of friction according to the use. For a brake member, facing or lining, it was suggested that the coefficient of friction range between .25 and .50 and for a clutch member, facing or lining, .10 to .35. Brinell hardness of the friction article ranged between .10 and .65, and "resistance to shear" was "sufficiently high to enable the articles to withstand the repeated longitudinal stresses ordinarily encountered in use." Certain tests for coefficient of friction, shear resistance, resistance to compression and wear were set up to which we shall refer later.
As to method, the specification stated that in general all ingredients should be in finely divided form with the lubricant and friction materials passing finer screens than the metals. The predominant metal in the composition should be finely divided and in the case of copper, the shape of the particles was important, "altho not absolutely essential." Fisher preferred to use "Koehler" copper powder whose particles were "flocculent or fern like and exhibited marked tendency to intertwine and matt together with large surface contact when subjected to pressure"; while stating that greatly improved results were achievable over any prior art, friction articles by using ordinary commercial nodular copper powder, the Koehler powder exhibited "a marked tendency to the formation of increased amounts of amorphous metal during pressing and to crystallization of such metal with a resultant network in which substantially all of the predominant metal particles are integrally united together into a continuous network resembling that of cold rolled and annealed brass or bronze." The specification disclosed an inverse ratio between strength, hardness and resistance to shear in the finished product and the specific gravity of the copper powder.
*508 Pressure was applied to the selected materials at approximately right angles to the wear face of the resultant article to obtain greater lubricating value from the graphite and mica content. Preferred pressures were from 10 to 18 tons per square inch, although lower or higher pressures could be employed and it was important that the article should be thin enough that the ingredients would be compressed uniformly throughout the thickness of the article which was to be in the shape of thin plates or bars.
After the pressing operation was completed, the article was subjected to heat treatment. "In general, the time and temperature of the heat treatment should be such that the lower melting point ingredients may melt and that certain of them, such as tin and zinc, may enter more or less completely into solution with the predominant metal, forming alloys * * * more or less uniformly scattered thruout the network." Temperatures from 1,000 to 1,300 degrees were suggested for periods up to 30 minutes. The heating should take place in an atmosphere substantially free from oxygen.
The specification provided that "Friction brake articles * * * may be produced in the form of thin disks, buttons or strips * * * with brads or rivets embedded therein." In the case of a clutch friction article, it was said to possess "a coefficient of friction which is low when there is slipping engagement with an opposed member, but which increases as the slipping decreases and as the engagement becomes closer or more positive and which reaches a maximum when engagement is complete. * * *." In conclusion it was stated that the term "friction article" included parts of clutches and brakes, but not bearings or bushings, "wherein the avoidance of friction is the object desired."
Claims 3 and 12 are typical of the article claims, and claim 19 illustrates the method. We quote 12 and 19:
"12. A pressed and heat treated, relatively dense, friction article suitable for lining clutches, brakes and the like, having a strong, malleable, metallic network substantially continuous, integral and free from interruptions and containing between about 65% and about 95% copper, between about 3% and about 15% of tin, and between about 5% and about 15% of lead, each 100 parts of said network carrying dispersed substantially uniformly therein between about 3 parts and about 20 parts by weight of graphite."
"19. The method of making friction articles of the character described, which includes the steps of mixing finely divided ingredients including between about 65% and about 95% of copper, between about 35% and about 5% of other metals such as tin, lead and zinc, and between about 3% and about 20% of inorganic non-metallic lubricant material and up to about 15% of inorganic non-metallic friction producing material, the mixture being substantially free from ingredients which may volatilize and form voids during the hereinafter described heat treatments, applying a pressure of between about 10 and about 18 tons per square inch to the mixture in a direction at right angles to the wear face of the resulting article to compress the mixture substantially uniformly thruout to a desired thickness, heating the pressed article between about 1,100 degrees and the melting point of copper for between about 3 and about 20 minutes in a protective atmosphere while preventing the formation of interruptions in the resulting metallic network in the article."
The "network" was a critical feature of the article claims. Finding of Fact No. 6 stated that the patent office was induced to issue the patent on the basis of affidavits which distinguished the alleged invention from the prior art by assertions that it contained a "metallic network" lacking in the prior art, but the court found (Findings 7 and 8) that the metallic network structure of friction articles was revealed in defendant's Exhibit 48, known as the Genelite block, which was made some 20 years earlier under the disclosures of Gilson No. 1,177,407; that it was also found in Williams No. 1,556,658 and in Claus No. 1,648,772. And the court also found (Finding 9) that "brake liners" using the same composition were sold by plaintiff to Cleveland Crane & Engineering Company as early as 1928, and as clutch brakes of "Genelite" *509 made under the Gilson patent to the Hudson Motor Car Company as early as 1920. It also found that the use of such composition for clutch liners was analogous to its use for brake liners; and in Finding 10 that the adaptation to a clutch liner of "friction articles" of the patent did not involve invention. In its Finding 5 the court stated that the processes employed had long been known to the old and highly developed science of metallurgy.
Under Rule 52 of the Rules of Civil Procedure, 28 U.S.C.A. following section 723c, we are not authorized to set aside these findings of fact unless they are clearly erroneous.
The plaintiff contends that these findings are refuted and negatived by certain statements found in the opinion of the court.
The court in its opinion [57 F. Supp. 220, 221], did say:
"Prior to 1932 there existed very serious clutch problems in the automotive industry. Such problems increased as the power and capacity of automobiles increased. The problem became quite acute about 1932 when the clutch linings then in use proved generally unsatisfactory for use in busses. The `conventional liners' at that time consisted principally of asbestos fiber in woven or molded form combined, with organic bonding materials. The friction to which these liners were subjected created such intense heat that it melted them. When their physical and operational properties were changed, their further use, if at all possible, became dangerous and unsatisfactory. * * *
"Many manufacturers were striving to devise a clutch lining that would slip smoothly when starting but would not slip when in driving engagement, that would not vary under varying conditions of pressure, temperature and speed, * * * and that would withstand the centrifugal, warping and twisting forces to which it was necessarily subjected.
"In 1932 the Twin Coach Company heard of plaintiff's `friction article'. It obtained some of the friction material and with plaintiff's cooperation installed segments as liners in clutches of its busses. Those clutch liners gave amazing results. They were subjected to severe acceleration tests and worked smoothly and proved to have remarkable durability. The evidence leaves no doubt that there had been a very great problem, that repeated efforts had come short of solution, and that the plaintiff's friction article supplied a long-felt need. Metallic clutch liners in a comparatively short time supplanted the `conventional liners' generally in the automotive industry."
We cannot agree with plaintiff's contention for two reasons, first, that the court itself did not of course regard any part of its opinion as a finding of fact; and second, that "amazing results" do not in and of themselves indicate invention. There must be "some substantial innovation * * *, an innovation for which society is truly indebted to the efforts of the patentee." (Italics ours.) Sinclair & Carroll Co., Inc. v. Inter-Chemical Corporation, 325 U.S. 327, 330, 65 S. Ct. 1143, 89 L. Ed. 1644. It is elementary that invention must involve something more than the work of a skilled mechanic charged with the state of the prior art.
On the question of the existence of the "metal network" in the prior art, there is evidence which supports the court's finding. Some of it is found in the specification of patent 070. The specification states that the term "network" was chosen as an expression more or less aptly describing the structure of the metallic body of the invention. The specification then went on to state that the parts might be likened to the rubber of a sponge in continuity and in the formation of space, and to the structure of sponge rubber wherein the rubber is more or less continuous and forms a network with many open spaces.
The analogy of the structural characteristics of a powdered metal product to the structure of the sponge is found in almost identical language in an article by E. G. Gilson of the Research Laboratory of General Electric Company, appearing in the publication, "Machinery" in October 1922, and in describing "Genelite, a New Bearing Material." He wrote:
"It is quite low in tensile strength, but comparatively high in compressive *510 strength. This will be readily understood when it is considered that the material is a bronze sponge with graphite distributed through its pores. The distribution of the graphite may be seen in the photomicrograph, Fig. 1, which shows the material greatly magnified. The black parts of the illustration represent the graphite, and the white the metal. The even distribution of the graphite and the continuity of the metal structure are well shown. The graphite is so firmly held that it is impossible to wash it out; in fact, it can be separated only by complete disintegration of the metal." (Italics ours.)
And the photomicrograph referred to clearly showed the continuous metallic structure as did the photomicrographs of 070.
The Gilson patent, No. 1,177,407, issued in 1916, for bearing material under which the Genelite material was manufactured, stated, "The present invention relates to bearing materials and comprises a coherent, spongy or porous body of metal having finely divided carbonaceous material, preferably graphite, distributed throughout its mass." (Italics ours.)
Obviously there was some continuity of metal both in plaintiff's structure and that of Gilson. Both used the analogy of the sponge and even though the plaintiff preferred the term "metallic network," the court was right in finding that the affidavits, which were instrumental in securing the issuance of the patent to the effect that the network did not exist in the prior art, were in error. Something more will be stated about the "network" feature later.
Plaintiff argues that even if the "network" existed in the prior art, the "bearing" field in which Genelite was used, was a different field from that of clutches and brakes, and that under the authority of Cincinnati Rubber Mfg. Co. v. Stowe-Woodward, Inc., 6 Cir., 111 F.2d 239, the citations were not anticipatory since they came from a remote and non-analogous art. We cannot allow ourselves to be drawn into the "semantics" of the distinction between a "friction" article, such as a clutch, and a "frictionless" article, such as a bearing.
A reasonable test is, whether the art from which the citations against the patent were drawn, was so remote that students of the questioned art would not naturally have looked there for help. We think that the "bearing" art was neither remote nor non-analogous. This is made clear in the final paragraph of the specifications of 070, wherein Fisher defined the term "friction article" as used in his patent by stating that it was "not intended to mean or include bearings, bushings and the like, wherein the avoidance of friction is the object desired." Manifestly, if the "bearing" art were remote and non-analogous, the necessity for making the distinction would not have occurred to Fisher.
See also the patent to Judy, No. 1,686,277, October 2, 1928, wherein he referred to metallic compositions of copper and lead, which were ideal for many purposes, such as "bearings, brake linings and clutch plates"; to Sherwood No. 1,873,223, August 23, 1923, wherein it was said that a process for manufacturing porous metal was applicable to the manufacture of "bearings, bushings, brushes, brake linings and other articles"; and to Perks, No. 1,919,168, July 18, 1933, wherein a composite friction material of asbestos and metal, or of combinations of powdered metals was said to be usable in "brakes, clutches, piston rings, etc." (Italics ours.) It is clear enough that bearings, brakes and clutch linings all lie within the same art.
Nevertheless, plaintiff contends that even if the "network" existed in the prior art, there was invention (1) in the new combination of the article claims; and (2) in the application thereof to a new use, that is, to clutch liners.
The second contention is met squarely by the references to Judy and Perks, wherein metallic combinations were considered for their use as clutch plates as well as for bearings and brake linings. We see no invention in applying the combination to a clutch use in view of the fact that Judy had already suggested the application of his metallic combination to clutches.
The contention that there was invention in the new combination of the article claims is closely allied with the question of *511 invention in the method claims and they will be considered together.
In 070 as already noted, the proportions of copper suggested were 65% to 95% by weight, of tin 3% to 15%, and of lead 5% to 15%, with certain substitutions indicated for tin and copper. Williams, No. 1,566,658, 1925, suggested for one use, copper 92%, tin 4% and lead 4%, all within the proportions used by Fisher. The proportions of graphite used in 070 were 5 to 15 parts by weight and in Williams 4% to 6%. In Gilson, No. 1,177,407, the proportions were copper 73%, tin 13% and lead 10%; and graphite 4%, all within the Fisher suggestion and supplying no basis for distinction. It is obvious from the ranges given by Fisher in 070, and stated in so many words as to the proportions of pore filler to be used, that adjustments in proportions were intended according to use. What the proportions were in specific instances were not given, not even in the claims; and experimentation for particular uses was clearly indicated. In Brady Brass Co. v. Ajax Metal Co., 3 Cir., 160 F. 84, 90, it was said: "A mere difference in the proportions of the constituents of an alloy, however useful the result may be, does not entitle the originator to the monopoly of a patent, in the absence of other circumstances than those here disclosed."
Plaintiff seeks to fortify Fisher's claim for invention on the ground that the copper used by him in his combination was not ordinary copper, but "strained copper," that is, copper which had been mechanically cold worked, or deposited by an electric current of high density, the flocculent powder of Koehler. In its brief plaintiff asserted that this copper contains stresses which during heating or sintering give it new and different properties; it crystallizes at lower temperatures, alloys more quickly with tin and forms stronger bonds with tin. "The `amazing results' are thus produced by a new combination of old elements one of which is new."
The brief continued,
"Fisher was the first one to devise a process which, in contrast with the prior processes of producing porous bearings, used a lower pressing pressure and a lower heating temperature for a short time. His new process gave `amazing results.' The quicker and more extensive alloying of the copper and tin resulted in the formation of alloy connections or firm bonds between the copper particles throughout the article and produced throughout the article a substantially continuous integral and uninterrupted metallic network which was strong, malleable and ductile; and the lower temperature, shorter heating, did not drive off, or `sweat out,' the lead but did melt the lead, so that it flowed around the network and graphite and anchored itself and the graphite in the network when the article cooled."
The point in these paragraphs is, that the beneficial results of the Fisher combination and process, go back to the unusual characteristics of strained copper. But the use of one form of strained copper, i. e., flocculent copper, in "bearings, contact members, and moving parts for engines, generators, motors," etc. was not new. See Koehler No. 1,714,564, 1929. But even if new, in the Fisher combination it is not claimed. If invention must stand on the use of strained copper, the court properly found that there was no invention since this critical element in the combination was not claimed. The patent grant can be construed to cover only what is both described and claimed. Miller v. Brass Co., 104 U.S. 350, 26 L. Ed. 783; Suczek v. General Motors Corporation, 6 Cir., 132 F.2d 371, 374.
There was nothing unusual about the process or method of mixing the ingredients of the Fisher friction article. This was done "in any convenient manner as will be understood by those skilled in the art." The pressure called for in Claim 19 was 10 to 18 tons per square inch, although the specification contemplated pressures above and below that range. Although method claims 18 and 19 stated that the pressure should be applied in a direction at right angles to the wear face of the resulting article to compress the mixture uniformly throughout, and in the specification it was stated that this provided from 2 to 5 times the lubricating effect since the flat surfaces were thus disposed parallel to *512 the wear surfaces, but Cover, plaintiff's expert, admitted that powdered materials had been pressed that way for a long time prior to Fisher. The pressure called for in the Williams patent No. 1,556,658, was 20-25 tons but Cover stated that he didn't think that the 20 tons was outside the 10 to 18 ton range called for by Fisher, though of course it was above it.
The specification called for heat treatment in temperatures of from 1,000 to 1,300 degrees for periods of up to 30 minutes. The two claims called for temperatures between 1,100 degrees and the melting point of copper for from 3 to 20 minutes. Williams called for 1,100 to 1,400 degrees for a period of from 5 to 7 hours. But plaintiff's expert Cover candidly stated with reference to the difference in the time of heat treatment, "as far as the results, I don't think it will make any particularly great difference," in the usability of the product. Fisher called for the heating to be done in a protective atmosphere, while Williams called for a non-oxidizing atmosphere.
Tests were conducted by the defendants with experts from both sides present. The tests were made upon compositions made respectively from certain prior patents, the Genelite block, and plaintiff's and defendants' commercial products. Test portions were mixed, pressed and sintered according to the specifications of the various patents; and then all test parts were subjected to uniform tests for strength, Brinell hardness, coefficient of friction and wear, according to the teachings of the Fisher patent. Defendants' expert, Wulff, testified that all materials met the specifications and claims of the Fisher patent. "There was a strength of over 200 pounds, Brinell hardness of above 10 and a coefficient of friction between .10 and .50." The wear tests were conducted according to the teachings of Fisher for brake materials. "The Claus materials, * * * showed a wear greater than .02 inches per hour. All others gave a wear of less than .02 inches per hour." The results of these tests seem convincing enough that the structural equivalent of the "network," if not its identical makeup, existed in earlier patents and materials.
We have examined but have not here reviewed all the evidence. Upon the best consideration that we can give, we cannot say as a matter of law that the court's findings of fact and conclusions of law as to the claims of 070 were "clearly erroneous." After all, Fisher's claims are improvement claims and while his article and method may have been a decided improvement over what had gone before, we cannot say that it represented more than what might have been expected of one reasonably skilled in the art. Adams v. Galion Iron Works & Mfg. Co., 6 Cir., 42 F.2d 395, 397; Stockham Pipe & Fittings Co. v. Ohio Steel Foundry Co., 6 Cir., 78 F.2d 111, 113.
Patent 460:
This patent was for an "article of manufacture" and obviously was framed for a practical application of 070. Plaintiff's brief described it as "a composite article which consists of a metal member, such as a steel clutch disk, integrally united to a liner such as that of the first Fisher patent. * * *" However, the specifications repeatedly stated that the liner of powdered metal might be adapted to "clutch plates, brake shoes, bearings and the like." (Italics ours.)
The specification states:
"It is an object of my invention therefore to produce improved articles such as bearings, clutch facings, brake linings and the like by firmly attaching material of the type referred to to metal backing plates either with or without the use of flux and either with or without simultaneous heat treatment of the powdered metal material.
"A further object of my invention is the provision of improved clutch plates, bearings, brake linings and the like composed of a metal backing plate having the requisite strength to which is bonded a facing of pressed powdered metal network material in which the step of heat treating the network material to obtain the desired characteristics thereof and the steps of bonding the facing to the backing are combined and carried out simultaneously." (Italics ours.)
The four claims in suit are article claims. Claim 1 is for a clutch plate; claim 3 for a brake element; and claim 5 specifies the *513 content and composition of the non-ferrous lining member. Claim 6 was as follows: "6. An article of manufacture of the type described consisting of a metal backing member and a non-ferrous lining member composed of powdered metals and friction and/or anti-friction materials formed into shape under pressure and heat treated to form a strong, malleable, metallic network substantially continuous and free from interruptions, said non-ferrous lining member being integrally attached to said backing member." (Italics ours.)
Each of the claims are alike in that in the article, integral attachment of the liner to the backing member was specified, or as claim 5 expressed it, "said non-ferrous lining member being bonded to said backing member at all points of engagement therewith"; and the specification stated, "In all of these articles it is extremely important that an absolutely integral structure be formed and that all possibility of separation be eliminated."
The court held that this patent was invalid for lack of novelty and invention and that it was anticipated by Short, No. 1,819,272, August 18, 1931. We think this conclusion is sustained. The Short patent related to "machine elements having surfaces adapted to be placed in rubbing contact with a relatively moveable part, such as a bearing for a shaft." The specification called attention to the fact that the metallic bond between porous and denser metals sometimes breaks loose when the machine element becomes hot. Short proposed to attach the porous member directly to the denser, reinforcing member, without the use of intermediate bonding material by "heating the two metal elements to a high temperature while clamped together under pressure."
Short also stated that when the porous, metallic facing was secured directly to the denser metal, better results were obtained if a portion only of the two contact surfaces were bonded, as by embossing and bonding at the point of raised contact, but he said "it will be understood that these methods of bonding are not limited to the use of the embossed element and may be employed to attach two plane surfaces together." (Italics ours.)
In view of what we have said in connection with our consideration of 070, we cannot say that Short was in a non-analogous art, and his patent so nearly anticipates the disclosures of 460, that we must regard 460 as invalid. Fisher must be held to have known of the Short patent and its disclosures. Adams v. Galion Iron Works & Mfg. Co., supra, 42 F.2d at page 397.
The decree of the District Court is affirmed.