Opinion
Nos. 9481, 9482.
April 16, 1948. Rehearing Denied May 14, 1948.
Appeals from the District Court of the United States for the Northern District of Indiana, Hammond Division; Luther M. Swygert, Judge.
Patent infringement action by the Linde Air Products Company against Graver Tank Manufacturing Company, Inc., and others. From the portion of judgment holding certain claims invalid, the plaintiff appeals, and from the portion of the judgment holding certain claims valid and infringed, the defendants appeal.
Affirmed in part and reversed in part, and cause remanded with directions.
Cahill, Gordon, Zachry Reindel, of New York City, Carlson, Pitzner, Hubbard Wolfe, of Chicago, Ill., and Crumpacker, May, Carlisle Beamer, of South Bend Ind., (Richard Russell Wolfe, of Chicago, John T. Cahill, James A. Fowler, Jr., and Loftus E. Becker, all of New York City, George M. Beamer, of South Bend. Ind., of counsel), for Linde Air Products Co.
Charles L. Byron, of Chicago Ill., John F. Oberlin, Ashley M. Van Duzer and James R. Stewart, all of Cleveland Ohio, and L.L. Bomberger, of Hammond Ind., for Graver Tank Mfg. Co.
Before SPARKS, MAJOR and MINTON, Circuit Judges.
Plaintiff charged defendants with infringement of United States Patent No. 2,043,960 issued June 9, 1936, to Union Carbide and Carbon Corporation of New York, assignee of the inventors Lloyd Theodore Jones, Harry Edward Kennedy and Maynard Arthur Rotermund of California, who filed the patent application on October 9, 1935. The patent relates both to a process or method of electric welding, and to compositions for use therein. Of the process or method claims, numbered 1 to 17 inclusive, all were in issue except claim No. 10. Of the composition claims, those numbered 18, 20, 22, 23, 24, 26 and 27 were in issue.
The defenses were invalidity and non-infringement. The court held valid and infringed composition claims 18, 20, 22 and 23. and invalid, composition claims numbered 24, 26 and 27. It also held all of the process or method claims in issue invalid. From the decree plaintiff, in cause No. 9481, has appealed from the rulings adverse to it, and the defendants, in cause No. 9482, have appealed from the rulings adverse to them.
The specification states with considerable particularity the disclosures of the art at that time, as patentees then understood them. We refer in substance to the specification in order to get a fair disclosure of the problems which then confronted the applicants, and which they sought to solve. The fact that the patent was issued clothes plaintiff with the presumption of its validity.
Among the various ways in which electrical energy had been converted into heat for welding metals, the arc process was the one most generally practiced, a typical example of which is its use to join the abutting edges of steel plates. In the metal-arc variation of the arc process, molten metal, provided by the melting of a metal wire or rod of suitable composition, was introduced between the abutting edges of the plates, and the latter were fused sufficiently to permit the added metal to coalesce with the metal of the plates so that, on cooling, a structurally strong bond resulted. The requisite heat was developed by maintaining either a direct current or an alternating arc between the parts to be welded connected to one side of the power line, and the wire or rod used to supply the molten metal connected to the other side of the power line. The abutting plate edges were usually beveled to form a trough for receiving the molten metal, thereby facilitating coalescence of the added metal with the plates throughout the thickness of the latter.
The simplicity of bare wire-electrode welding recommended it above all other methods, but the metallurgical and physical properties of metal deposited in this manner were usually so poor that the method was unsuitable for many applications.
In order to avoid certain difficulties encountered in bare metal arc welding, it was customary to protect the freshly deposited molten metal with a blanket of molten metal compounds (usually compounds of the alkali or alkaline earth metals). The material used for forming this blanket is called the "flux." The usual method of providing a flux blanket on the weld was to encase the welding rod or wire, referred to as the electrode, in an adherent sheath of solid flux, and this gave rise to other difficulties. The sheath was usually fragile and, being nonmetallic in nature, was non-conductive when cold, so that electrical connection was required to be made with the electrode at points bared for that purpose at intervals and at variable distances from the weld, thereby imposing an additional burden on the automatic regulating devices widely used in welding. When the bared sections reached the melting zone, metal of inferior quality was deposited. The current passed through a variable length of rod, the length between the arc and the contact point, heating it sufficiently to crack off the flux covering and further to add an IR-drop to the arc voltage. This IR-drop was not constant in magnitude but varied according to the position of the contact point relative to the arc. The machine could not maintain the constant arc length so necessary to successful welding unless compensatory means were resorted to. These difficulties therefore limited the energy which might be expended in the arc. The inventors were not aware of current values in excess of about 500 amperes being used in this manner of welding.
For good welding a homogeneous deposit of metal is indispensable. In metal-arc welding the difficulties in securing such deposit are multiplied as the thickness of the deposit increases so that, in such case, it becomes necessary to make several traverses or passes, adding a layer of metal each time until the necessary thickness is attained. Such method is time and labor consuming.
Another known means of applying protective flux consisted in utilizing a thick coating of finely divided material, a flux, which covered the seam to be welded. The welding electrode which is a bare wire of suitable composition was fed down by conventional feeding means and the arc was struck under the powdered flux. The electrode was traversed at a constant rate along the path to be welded. Direct current was applied at a voltage of about 30 volts with a current up to about 900 amperes. The flux used in that system was a natural clay of approximately the composition used in making brick. Although this method retained the advantage of feeding current continuously to a bare wire at a point adjacent the weld, the weld metal thus deposited thereby was deficient in quality. A weld thus made had ample strength for some purposes, but it was quite porous, was not applicable to plates over one-half inch thick, and projected a cloud of material into the atmosphere, and often required the welders to use gas masks for protection from the dust.
The objects of the present invention are said to be to avoid the disadvantageous features above referred to; to provide a process by which thick plates can be strongly and rapidly electrically welded in a single pass or in a plurality of passes, with a weld of which the density and physical properties are at least equal to those of the parent metal; to provide a process in which heavier welding currents can be used and the rate of welding accelerated; to overcome the difficulties which, in prior processes, are caused by the inherent instability of an electric arc, and its liability to be extinguished by a variety of causes; and to avoid the necessity for a flux sheath on the electrode.
It is said that the patent discloses a novel process for electric welding wherein the necessary heat is generated by the passage of a heavy electric current between a metal electrode (usually bare) and the metal plates or similar objects to be welded, the electrode being out of mechanical contact with the objects, and the current being carried across the gap between the electrode and the object by and through a conductive melt or welding composition having appropriate electrical resistance properties. The heat thus generated melts successive portions of the electrode, and the molten material is deposited as weld filler material. The welding composition serves as an active instrumentality or welding medium, inasmuch as it provides heating means, controls the rate, penetration and quality of welding, and purifies and protects the molten metal.
The specification states that the properties of a successful welding composition for this method are: (1) The chemical reaction between the components of the welding composition must be completed before it is used in welding (failure in this regard produces porosity); (2) it must be capable of controlling the penetration and the width of the weld; (3) its fluidity at welding temperatures must be such that it will not become entrained with the molten metal; (4) it must consist of chemicals which are not detrimental to the properties of the steel; and (5) it must be readily removable from the finished weld.
The composition of the welding medium is of the utmost importance. The particular composition that is to be used is determined by the quality and thickness of the metal plates to be welded, by the current and voltage to be used, and by the properties it is desired to impart to the weld metal.
The inventors state in their specification that they have used calcium silicate and silicates of sodium, barium, iron, maganese, cobalt, magnesium, nickel and aluminum, both in binary and ternary combinations, in various proportions. They state also that they have used calcium titanate and various titano-silicates, particularly when they desired to introduce titanium into the weld metal. They further state that a number of these conductive welding compositions are more or less efficacious in this process, but it is preferred to use silicates of the alkaline earth metals, such as calcium silicate, and they also prefer to add to these silicates minor proportions of alumina and of a substance adapted to lower the melting point, such as a halide salt.
The preferred composition, however, comprises, as its principal ingredients, silica, at least one basic constituent consisting of an alkaline earth such as lime or magnesia or a mixture thereof, and alumina.
The silica and basic constituent are in approximately the proportions theoretically required to form metasilicates, although a substantial excess of silica is permissible and frequently advisable. For example, if the basic constituent is lime, the proportions by weight would be approximately from 0.7 to one part of lime to one part of silica; and if magnesia is substituted for part or all of the lime, the silica is increased accordingly, so as to maintain about the same stoichiometric proportions. The alumina forms about 1% to 8% of the weight of the welding composition, and preferably about 4% to 6%. In order to cause the energy to be distributed over a larger area there is added, in various minor proportions, say 6%, a material capable of reducing the energy concentration and thereby widening the weld. The preferred material for this purpose is a halide salt such as calcium fluoride.
The basic constituent preferably consists chiefly of lime and magnesia in approximately the proportions of three parts of lime to one part of magnesia, by weight, but other proportions are useful and within the invention. Basic materials other than lime or magnesia, for example oxides of manganese, titanium, or the alkali metals, and known fluxing agents such as borax or boric acid, may be added in moderate amounts without greatly changing the essential characteristics of the composition, but it is important that the composition be substantially free from iron oxides uncombined with other ingredients of the composition (by being prefused with such ingredients) and from materials (such as carbonates or moisture) which evolve detrimental amounts of gas or vapor at welding temperatures. The material when ready for use to form the conductive melt is said to be the principal feature of the process.
The chemical condition of the welding composition is important. "The acidic and basic constituents should be reacted, the composition should be substantially anhydrous and free from gases, and all reactions which would evolve deleterious amounts of gases during welding should be substantially completed, before the medium is used in the welding process." These conditions are attained by prefusing a mixture of the constituents. Certain relatively volatile but chemically stable constituents, such as calcium fluoride, may be added to the medium either before or after the other ingredients have been fused and cooled. The carbonates of calcium or magnesium may be substituted for the oxides, if the ingredients are prefused.
It is also said that the physical condition of the welding composition is of the utmost importance. If the molten medium is rapidly solidified by cooling before grinding and use, more homogeneous and solid welds may be produced than when it slowly loses heat and solidifies. The fused composition is preferably cast as a relatively thin layer against efficient cooling or chilling means such as a cold, heavy plate or block of metal, or a water-cooled chill plate, rather than as a large ingot. It is preferred that the rate of cooling be such that substantially all of the solidified welding composition is characterized by a vitreous fracture.
In practicing the process the plates to be welded are arranged as for metal open arc welding. The edges are preferably beveled, but this is less necessary than with arc welding because the "penetration" and width of the weld is dictated by the composition of the flux and the voltage and current employed, and can be closely controlled by the process.
When a weld is to be made between beveled edges, the operation is started by filling the groove, space, or gap between the edges of the objects with the welding composition, preferably in powdered or granular form, and a substantial and additional quantity of the composition is heaped up along the groove, thereby overfilling it. After the weld is completed, that portion of the composition which has not been fused may be removed, and that portion which has been fused may be lifted from the weld. The fused material may be reground and reused in some instances.
Pursuant to the method prescribed, the end of the welding electrode, which is preferably bare metal, is first inserted into the welding composition. Since the latter is non-conductive when cold, a conductive path for the welding current is provided by bridging the gap between the electrode and the work with a sliver of graphite or a wad of steel wool. The power is then applied, the welding composition is locally heated until it fuses and becomes conductive, forming a subsurface pool, and immediately thereafter the end of the electrode begins to fuse and the molten metal begins to deposit in the groove, displacing the subsurface pool of fused welding composition. At the same time, the edges of the object being welded begin to fuse and to coalesce with the deposited metal. The electrode is then fed toward the work and moved along the seam at an appropriate rate preferably by a mechanical device which may be similar to those which are extensively used in open arc welding, and the welding proceeds without interruption under a protective layer of the composition. There is practically no puffing or visible displacement of the welding medium, no visible sparking and scarcely any external evidence that the weld is being made. The weld is ordinarily completed in one pass, even in thick plate. The welding composition in front of the electrode remains unfused, while the fused composition rises to the top of the completed portion of the weld and solidifies. This effect is scarcely visible because of the use of an excess of the welding composition which remains unfused and covers the fused composition and metal.
Either alternating or direct current, or direct current superimposed on alternating current may be used, but alternating current is preferred.
The application states that the process here employed does not depend on the formation of an arc of the usual type:
"* * * If, while a weld is being made, the circuit is opened externally of the weld and then closed again while the resistive composition is still molten, the current will immediately resume its flow without any necessity for moving the electrode into contact with the seam as it would be necessary to do to reestablish an arc of the usual type. The heat for melting the electrode is evidently developed in the conductice melt itself.
"The conductive welding composition used in our process doubtless performs the protective functions of a flux, but it is much more than a flux, and many compositions which are satisfactory fluxes for arc welding are unsuitable for our purpose. * * *"
The District Court said, and we think rightly, that Jones, Kennedy and Rotermund invented something patentable over the prior art of electric welding. He thought their method of welding differed vastly from electric-arc hand welding, either of the bare or coated rod variety, and that there was "an inventive level-difference" between their method and Robinoff's automatic welding method. He calls attention to the differences in the physical appearances in the welding zones of the patent, and those of any of the prior art during their operations. "In bare rod and coated rod hand welding, the electric arc creates a blinding glare and there is a great amount of smoke and splatter. In the Robinoff clay flux method constant flashings occur, and the arc is always at least partially visible." In those of the patent "there is no glare, no open arc, no splatter, and very little, if any smoke, * * *. The truly remarkable difference, however, between what Jones, Kennedy and Rotermund invented and what had gone before is perhaps best manifested by the performance achievements of their invention. For instance, only through its use can plates as thick as two and one-half inches be welded in a single pass. Furthermore, the welding speeds made possible by it dwarf those of any other method, and the welds produced by it are of the highest quality in contrast to the great amount of porosity contained in the welds produced by the so-called clay flux process."
Having thus determined that there was patentable invention, the court then proceeded to determine what was patentable, whether process or composition or both. As to the method he posed the question "whether a fundamental difference in phenomena exists between the Jones, et al. method and all methods of electric arc welding. If it could be said without equivocation that the electric current passing between the electrode and the base metal travels through a welding composition which is in a melted or liquid state and that no 'electric arc' is present, then a different electrical phenomenon would exist and a radically new process in electric welding would have been discovered. But no such finding can be made. The evidence is persuasive that no such basic difference in phenomena is present in the Jones, et al. method."
We construe this statement to mean that the court thought the invention as to method, would have been patentable if it had disclosed that there was no arc present, and that the electrical current passing between the electrode and the base metal traveled through the welding composition, only when such composition was in a melted or liquid state.
With respect to this point, the court in substance said that in the inventors' experiments it was hoped by them to give a better weld than was being obtained by the Robinoff process, which involved an electric arc phenomenon. During the course of their experiments, they evolved a composition which produced better welds. However, he said in substance that the purely operational use of their welding composition was the same as the clay flux in the Robinoff process, and there was and is no difference between the Robinoff and Jones and other methods. In both methods a deep layer of flux or welding composition is laid in the V formed by the beveled edges of the plates to be welded; a bare welding rod is fed down through this composition; and an electric current which produces the heat to weld the plates is conducted from the rod to the base metal. Thus, the court said, the physical operational steps in the two methods are identical.
Plaintiff's appeal is taken from so much of the decree as held invalid claims 1 through 9, and 11 through 17, and claims 24, 26 and 27. The court considered these claims in four groups, and as a matter of convenience we shall do likewise.
Claims 1, 3, 4, 7, 8 and 9 were held to be invalid because they were too broad and indefinite to satisfy the statute, that is to say, because they were said to make no reference to the chemical constitutents of the welding composition. These were not claims to a composition of matter or chemical compound but they were all process or method claims as expressed in finding number 7: "Their invention also included a novel welding process in the sense that the electric arc present between the electrode and base metal is sheathed by a mantle of molten welding composition and in that the conductive arc area is composed predominantly of vaporous welding composition." There the court correctly described the alleged novel process without in any way referring to any chemical composition.
We think it was unnecessary in a process claim to refer in any way to the chemical constituents of the composition where, as here, they are fully set forth in the specification. We think this view is fully supported in principle by the Telephone Cases, (Dolbear v. American Bell Telephone Co.), 126 U.S. 1, 8 S.Ct. 778, 31 L.Ed. 863; Union Oil Co. of California v. American Bitumuls Co., 9 Cir., 109 F.2d 140; Van der Grinten v. Dieterich-Post Co., 9 Cir., 85 F.2d 637; and Glade v. Walgreen Co., 7 Cir., 122 F.2d 306.
Furthermore, these claims now under consideration seem to be sufficiently definite when construed in the light of the specification. The statute merely states that the claims should particularly point out and distinctly claim the part, improvement, or combination which the applicant claims is his invention or discovery. 35 U.S.C.A. § 33. It would seem that this statute only requires that the claims point out the invention, not that they redescribe it.
The court's reason for holding invalid process claims 5 and 6, and 11 through 17, is set forth. "* * * The import of each of these claims is that the sole conductive medium through which the electric current passes from the electrode to the base metal is the welding composition which is in a molten state and that the electric arc phenomenon is not present. * * * While a patent covering a meritorious invention should not be struck down because the patentee has misconceived the scientific principle of his invention, the error cannot be overlooked when the misconception is embodied in the claim." (Our emphasis.)
"The process of electric welding which comprises the steps of juxtaposing a metal electrode and a conductive member on which metal is to be deposited; establishing between said member and said electrode a conductive melt comprising an alkaline earth metal silicate, alumina, and an alkaline earth metal fluoride, substantially free from substances, such as uncombined iron oxide, capable of evolving gases when fused in contact with the metal to be welded; passing through said melt an electric current of sufficient magnitude to maintain the conductivity, progressively melt the said electrode and cause the resulting molten metal to coalesce with said member; and maintaining said melt in contact with said electrode and member throughout the welding operation."
None of these nine claims refer in any manner to the presence or absence of an arc, either directly or by reasonable inference. The reason is that the applicants did not know, and they carefully avoided any commitment on that subject. There was no visible evidence of its presence, and for that reason it was impossible for them to determine that scientific fact, nor was it incumbent upon them to do so. The file wrapper discloses that the applicants said to the Examiner:
"It is difficult, if not impossible, to define an arc precisely, and applicants have contemplated the possibility that a definition may be framed in such terms as to embrace the phenomenon which occurs in their process.
"There can be no question, however, that there is no such arc involved in their process as occurs in every prior art process including that of Robinoff, et al."
We are convinced that the District Court also erred in its statement, which we have quoted, to the effect that the import of these claims is that the sole conductive medium through which the electric current passes to the base metal is the welding composition which is in a molten state. This no doubt occurred through its interpretation of the phrase "conductive melt," coined by the inventors to identify their welding compositions. This interpretation, if correct, would preclude the existence of such compositions in vaporous form. It is clear that in this respect the court should adopt the same interpretation of the term "conductive melt" which the applicants reasonably expected it to have when they coined it. It is clear to us from its context in the patent, and the circumstances of its origin clearly indicate that the applicants intended the phrase to mean, when heated to welding temperatures, in either liquid or vaporous state.
This construction is quite consistent with the views expressed by experts of both plaintiff and defendant, with the Welding Handbook of the American Welding Society, and with Webster's New International Dictionary since 1936, which defines "fluid" to mean "liquid or gaseous." The specification expressly states "that the relatively high current densities which are preferably used superheat the molten metal and welding composition, making them highly fluid."
The District Court held composition claim 27 invalid because it contained the words "a prefused electric welding composition which is conductive when molten * * *." The court, as before, thought that these words gave the inevitable impression that the electric current passes through the composition only while it is in a molten state. From what we have just said with respect to the last method claims, we think the holding was erroneous, and the same may be said with respect to process claim 1.
The District Court held process claim 2 and composition claims 24 and 26 invalid on the ground that a number of silicates had been shown to be inoperative. The composition is described as a silicate in process claim 2, as a metallic silicate and calcium fluoride in composition claim 24, and as consisting chiefly of silicates in composition claim 26. It rejected plaintiff's contention that "silicates" and "metallic, silicates," when read in the light of the specifications, meant the nine metallic silicates which had been proved operative. The court based its ruling upon Aluminum Company of America v. Thompson Products, Inc., 6 Cir., 122 F.2d 796. We fully agree with what is there said. However, there is nothing in this record to support the fact that the applicants for this patent intended by these claims to assert any monopoly broader than nine metallic silicates named in the specifications. Under these circumstances, we think the court should have applied the general rule as laid down in the case just referred to and followed the construction which would sustain them.
It is only when these claims last referred to are read alone, without regard to the rest of the patent, that the terms "silicates" and "metallic silicates" can be construed as meaning all "silicates" or all "metallic silicates." It seems clear to us that when the patent is read as a whole it is quite plain that these terms refer to the three outlined earth silicates plus the six other metallic silicates listed in the specifications as their equivalents. We think the court should have held valid the process claim 2 and composition claims 24, 26 and 27.
The only issues involved in defendants' appeal are the validity and infringement of composition claims 18, 20, 22 and 23. The latter is typical.
Claim 23. "A finely-divided unbonded fusible composition for use in electric welding employing a bare metal electrode as a source of weld metal, said composition being substantially free from substances capable of evolving a detrimental quantity of gas under welding conditions and containing calcium fluoride and a major proportion of one or more alkaline earth metal silicates."
Plaintiff's process was quite different from that of the prior art disclosures in that the latter depended on the evolution of more gas to protect the molten metal, while plaintiff's proceeds on the theory of eliminating all gas forming ingredients. That difference is clearly one of process. True, plaintiff's process contemplated and depended upon a different composition of flux, and the patent was issued on both process and flux composition, and we think rightly so. There are as many different kinds of flux as there are elements and combinations of elements, and when thus used their results are quite as numerous. However, this patent was not granted upon a result, but it was granted upon a process or method which produced a much desired result, and which was then unknown to the art. Any other process or method which produces the same result would not infringe this process patent, provided it does not use this method or its equivalent, and this would be true regardless of what flux, if any, was used for such purpose.
The court found that the patent included a novel welding process in the sense that the electric arc between the electrode and base metal is sheathed by a mantle of welding composition and that the conductive arc area is composed predominantly of vaporous welding composition; that the capabilities of the invention represented an important advance in the welding art; that compositions of calcium, magnesium, strontium and barium, when prepared in accordance with the specification, are operative as welding compositions within the teachings of the patent; that these elements are known in chemical literature as alkaline earth metals; that manganese is not an alkaline earth metal; that manganese silicate is one of the nine metallic silicates which the specification discloses that the inventors have used as principal constituents in welding compositions attaining more or less effectively the objectives of their invention; that manganese silicate when used as a principal constituent of a welding composition, prepared according to the teachings of the patent, is a proper substitute for the silicates of the alkaline earth metals.
The court further found that the defendants used and sold their accused composition since 1942, and are still using it, and it is substantially identical in operation and result with that of the patent; and that such flux is a calcium manganese aluminum silicate, with calcium fluoride added, which in welding operation is in all respects equivalent to the compositions of the patent; that a person skilled in the art can produce the accused flux of the defendants by following the teaching of the patent in suit; that no evidence was introduced to show that the accused flux was derived from the prior art or by independent experiment, or from any source other than the teaching of the patent in suit. It was further found that in the defendants' flux the slight excess of basic oxide over stoichiometric proportions does not substantially affect its operative character and it follows the teaching of the patent; that the patent teaches the substitution of manganese for other basic oxides and is not limited to the use of a moderate amount of manganese in addition to other basic oxides.
The court further found that plaintiff does not, as alleged by the defendants, impose as a condition of the license under the patent in suit, nor does it ever impose on its licensees any requirement, condition, agreement or understanding as to the purchase or use of merchandise, machinery or supplies or other commodities; and the licensees are free to buy and use any materials and equipment from any source. These questions of fact are supported by substantial evidence and we may not disturb them.
We find no error in the court's ruling on the issues raised by the cross-appeal. We affirm so much of the District Court's decree as holds valid and infringed claims 18, 20, 22 and 23. We reverse that part of his decree which holds invalid all of the process or method claims in issue, and composition claims 24, 26 and 27. The cause is ordered remanded to the District Court for further proceedings not inconsistent with this opinion.