Opinion
No. 7521.
Heard May 5, 1970.
Decided May 18, 1970.
Herbert P. Kenway, Boston, Mass., with whom Robert J. Horn, Jr. and Kenway, Jenney Hildreth, Boston, Mass., were on the brief, for appellant.
Ira Milton Jones, Milwaukee, Wis., with whom James R. Custin, Milwaukee, Wis., and Robert L. Thompson, Boston, Mass., were on the brief, for appellee.
Before ALDRICH, Chief Judge, McENTEE and COFFIN, Circuit Judges.
This is an action for infringement of Anderson Patent No. 3,315,548, dated April 25, 1967, now Re. 26,676. Plaintiff, owner, is Contour Saws, Inc.; defendant, The L.S. Starrett Company. The patent is for a method of making a bimetal band saw blade having a low alloy base with tooth tips of high speed tool steel. The principal defense is invalidity. The district court found the patent valid and infringed, and defendant appeals.
A metal-cutting band saw blade must have great flexibility and resistance to fatigue, and hard cutting teeth. The two require characteristics sufficiently different to make a compromise inevitable, in spite of even complex treatment, in a blade of a single metal. Early attempts at producing a bimetal blade were either inefficient, or impractically expensive. An example of the latter was Wilkie, No. 2,318,549, by which a piece of hard material was brazed to the leading edge of each tooth. Other methods were found that were less expensive, but none was satisfactory, chiefly because of difficulties arising out of the welding.
In the late 1950's there was developed in Germany a method of welding by an electron beam. On October 21, 1963 there issued a Belgian patent, No. 629,665, disclosing the application of this process to produce bimetal saw blades. In November 1964 plaintiff introduced an electron welded band saw blade on the market, and the following month filed an application which covered not only electron beam welding, but a heat treatment process which hardened the teeth and yet preserved the required characteristics of the back-up metal. This resulted in the patent in suit. Defendant immediately copied it. The parties have stipulated that the process described in the patent is now widely used by the plaintiff and its licensees, and is the only one that has been commercially successful.
Plaintiff's position on this appeal is that no questions of law are involved and that the only issue is whether the district court's findings are plainly wrong. Defendant does not seriously contradict this, but contends that the court was in fact plainly wrong on the issue of validity in not finding obviousness, or, alternatively, that the patent was "substantially anticipated" by, or obvious in the light of, the Belgian patent aforementioned, and German Patent No. 974, 524, dealing with a heat process. Its claim that the court was plainly wrong in finding infringement lacks all merit, and need not detain us.
Defendant says that it does not use a "muffle" furnace, a term used in the patent. This is based upon the testimony of its expert as to what, assertedly, is understood in the art by this term. The patent described its furnace by a specific reference to another patent, in which its design is fully explained. Defendant uses precisely that furnace. It is of no conceivable consequence that plaintiff's patent may have misused the word "muffle."
The court dismissed obviousness apart from the foreign patents almost by default, and correctly so. Defendant, a substantial competitor of the plaintiff, and highly skilled in the art, learned of plaintiff's product in June 1965. It was at once apprehensive of what it described as a "severe" threat to this branch of its business, and its test reports of plaintiff's product spoke only in the highest terms. Yet, when the patent issued, nearly two years later, defendant did not have even shop production. Even an attempt to make something less good was not persisted in. Three months after the patent issued, defendant was in production.
The total process had been just as unobvious to the plaintiff as it was to the defendant. The unforeseen effect related to the heating procedure, which worked out successfully when the plaintiff had expected the contrary. The process is this. In order to cut through steel and even harder metals the teeth of the saw must also be hard. Furthermore, as a matter of economics, the teeth must be able to be formed by milling, rather than by grinding, which means that this must be accomplished after the blade is otherwise ready, but before the final hardening process. The grinding is then confined to the sharpening. The back-up of the teeth, the body of the blade, on the other hand, cannot be this hard, or it will resist bending. This is not serious in a hack saw blade, except that hard steel is more expensive, because a hack saw reciprocates in a single plane. A band saw, however, continuously passes over a top and bottom wheel, with straight runs in between. For controlled cutting, the run must be at a high tension. Hence the blade, operating at a fast rate of speed, is constantly and forcefully flexed and unflexed.
The steel that will best resist fatigue by flexing must be relatively soft in composition, and is known as spring steel. To best resist abrasion, the steel for the teeth should be hard in composition, commonly known as high speed, or tool steel. In addition, as is well known, the heat treatments to attain such results differ substantially. When the blade was made out of a single piece of metal various methods were developed to give a different heat treatment to the tooth edge or to the teeth, and to the blade. Also, different compromises were made as to the blade's composition. However, the obvious desideratum was a bimetal blade. This was first achieved by resistance welding, the faults of which occupy much of the present record.
Electron beam welding proved to have no faults. The matters of which plaintiff would make much are nothing but ordinary troubles not unusual in the mechanics of setting a theory to practice. There is no evidence that plaintiff made any significant contribution with regard to the welding disclosure in its patent. Where the process was already known, even had plaintiff been the first to apply it to saw blades (which, in the light of the Belgian patent, infra, it was not) it is doubtful whether this would have been invention. Warden v. City of St. Louis, Mo., 8 Cir., 1944, 140 F.2d 615, cert. denied 322 U.S. 746, 64 S.Ct. 1155, 88 L.Ed. 1578; Allen v. Coe, 77 U.S.App.D.C. 324, 1943, 135 F.2d 11. The "discovery" of efficiency is not inventive. General Electric Co. v. Jewel Co., 1945, 326 U.S. 242, 66 S.Ct. 81, 90 L.Ed. 43.
What proved to work out in an unexpectedly successful fashion was the heat treatment afforded by plaintiff to the electron welded blade. After the tool steel strip has been welded to the blade strip, and the teeth cut and set, to harden them necessitates the application of high heat. The patent recited 2150° to 2275°, which is concededly standard. However, for the spring steel blade, or base, to be subjected to such a temperature would, according to accepted knowledge, be expected so to coarsen the grain and degrade the metal that it would lose its fatigue resistance quality. After tool steel has been hardened it must be annealed or tempered. One claim of the patent gave no tempering temperature; the other said 1050°. Plaintiff discovered that the tempering rectified the damage done by the previous overheating.
Plaintiff's claim in its brief of an alleged synergistic effect of the combination of electron welding and heat may be readily disposed of. It is without any findings of the court, and without a word of evidentiary support. Even had plaintiff established some small degree of benefit in the combination, this would not have constituted invention. See United Shoe Mach. Corp. v. Industrial Shoe Mach. Corp., 1 Cir., 1964, 335 F.2d 577, 580, cert. denied 379 U.S. 990, 85 S.Ct. 702, 13 L.Ed.2d 610.
On the record, this was an achievement. It is true that defendant in 1964, for experimental purposes, made a bimetal band saw welded by a high frequency induction or resistance method. It asserts that it used this same heating and annealing procedure. There is no very compelling evidence that it did, and none at all that defendant discovered the properties announced by the plaintiff. On the contrary, all we have is a report that the blade continually broke after 18 hours of use.
In view of this, plaintiff's case must succeed unless it founders on the Belgian and German patents. The court found that while the former discloses "the use of an electron beam for joining materials in the making of saw blades; * * * it does not show how electron beam welding might be employed in the making of bimetal band saw blades, and it does not specify the particular metals to be joined in making the saw blades to which it does relate, nor how to treat them." This finding, with a few minor grammatical changes, was an exact adoption of plaintiff's requests for findings. Except for the last five words, it is totally incorrect.
We have already noted the significant difficulties attached to the production of band saw as distinguished from hack saw blades. Plaintiff's unfortunately successful misguiding of the court thus denigrated the Belgian patent. But in stating that it did not apply to band saw blades plaintiff spoke, and the court found, in the face of no less than five references in the patent itself. We need quote but two. We do not understand how the court was misled, but even less do we understand how the plaintiff could have made such a request. When we inquired at oral argument, its counsel conceded that there was no testimony supporting this conclusion, and said it was his own opinion. No reading of the patent can justify it. As counsel's ipse dixit it is valueless.
"The invention concerns metallic saw blades such as circular saw blades, band saw blades. * * *
"It is already known in the prior art to make circular saw blades. * * * However, it has not been possible heretofore to apply this process to band saw blades. * * *"
Similarly, the court's finding, at the request of the plaintiff, that the Belgian patent "does not specify the particular metals to be joined" is contrary both to the language of the patent, and to general understanding in the art. The patent stated, "Saw blades have previously been constructed of two different materials: the body of the blade or the back portion * * * being made of material of low quality, while the front part of the blade, which comprises the teeth, must be made of material of very high quality, such as for example tempered steel." The patent went on to state that these two materials could now be joined by the electron process. It is true that the patent did not specify the exact composition of the low quality blade or of the high quality teeth, but if the criticism of the patent is to be read in that sense, it is meaningless. The patent is significant precisely in that it is of general application.
In addition to misdirecting the district court, the plaintiff goes even further with us. In its brief it says, "The Belgian patent does not suggest welding a strip of high speed steel to one edge of a low alloy backing band." (Emphasis in orig.) "The Belgian patent teaches a tooth-by-tooth application of hard metal tooth elements to a blade backing. * * * The teachings * * * are those of the old Wilkie patent." (See, as to Wilkie, paragraph 2, supra, of this opinion.) The patent itself, however, states something quite different. "The present invention * * * provides a saw blade made of two rectangular parts, one of which is of a material of high quality and the other of a material of lower quality, which are welded electronically, and which are subsequently shaped in such a way that the rear portion of the blade, that is to say, the portion that carries the teeth. * * *" Nothing is said about welding teeth, which would be a multitude of parts. The "two rectangular parts" of different metals are plainly the two strips that plaintiff's counsel, again without benefit of testimony, says they are not.
It is true, as the court found, that the Belgian patent does not teach the heat treatment to be given to the welded blade. Here defendant points to the German patent, No. 974,524, dated January 5, 1961. This patent recited that when bimetal saws, consisting of a "carrier body" of ordinary steel having limited hardening capacity with low tensile strength, and a cutting edge of high-speed steel, were subjected to the required hardening temperature of high-speed steel, "the carrier body steel reaches the region of the overheating temperature." The patentee has discovered that the subsequent "tempering of the high-speed steel closely below 600° C." improves the condition of the carrier body. "The damages of overheating appear to have been fully compensated." This is exactly the "discovery" that plaintiff made three years later.
In its brief plaintiff says that there is not a "scintilla of evidence in the record that the tempering temperature specified in the patent in suit (1050° F. * * *) is "closely below' the temperature recommended in the German patent." Apart from the fact that one claim of plaintiff's patent specifies no tempering temperature, 1050° F. is 583° C. If 583° C. is not "closely below 600° C." we need more than counsel's say-so.
Plaintiff's response to this is two-pronged. The first, reminiscent of its attack on the Belgian patent, is to claim that the German patent did not relate to band saw blades. Responding to plaintiff's requests, the court found that this patent, too, related to "hack saw blades." It is true that in this instance the patent did not mention band saw blades specifically. Defendant's expert, however, testified that band saw blades were included because the reference was to "a metal cutting saw blade submitted to severe bending." This, of course, is just the difference between band saw blades and hack saw blades. No one contradicted the witness' seemingly obvious conclusion, save plaintiff's counsel. The court's finding is unsupported.
Plaintiff volunteered in oral argument that the court did make one, "minor" mistake. This was its finding that the German patent "contains no disclosure of a 2150°-2275° F. heat treating temperature." Plaintiff's counsel suggested that the court overlooked the significance of the fact, brought out elsewhere in the record, that there is an understood standard range of 2150°-2275° for heating high speed steel, which means that there was no need for the patent to mention it. It is true that the court mistakenly overlooked this fact. But the court's primary mistake was making, in haec verba, the finding of the patent's failure to specify a temperature of 2150°-2275° that the plaintiff asked it to make.
We observe in the record the place where the court lost confidence in one of the defendant's witnesses. This was no reason to place total confidence in plaintiff's counsel. We have had occasion to comment upon the sometime necessity in difficult patent cases of the court's reliance upon counsel's requests for findings when technical complexities and the court's lack of expertise prevent full independent analysis of the record on its part. Nyyssonen v. Bendix, 1 Cir., 1965, 342 F.2d 531, cert. denied 382 U.S. 847, 86 S.Ct. 63, 15 L.Ed.2d 86. In the case at bar, which the plaintiff considered so simple that it called no expert witness, there was no such obstacle.
Nor was this a "minor" mistake. The supposed significance of this finding was that plaintiff's patent gave a temperature range of 2150°-2275° and the German patent, by giving none, did not correspond. In fact, it corresponded exactly. But if the court could be led to think that it did not — and that, as previously mentioned, it did not even relate to band saw blades — the German patent could be given a quiet burial.
The most charitable interpretation that we can give to plaintiff's repeated misdescription of the Belgian and German patents is that its fears of their destructive effect upon its patent have affected its judgment. However that may be, we have gone into this lengthy analysis not so much because of our dislike of unjustifiable statements from parties and counsel, but to demonstrate the sole, and erroneous, basis of the findings of the district court. So many of them are plainly wrong that we cannot accept its ultimate conclusions.
The question comes, what do we do. It seems to us, on reviewing the record, that we have a clearly dictated answer. There was no invention in adopting the electron beam welding process to the making of bimetal band saw blades. There was no invention in learning that the weld, advertised by the makers of the electron machine which plaintiff used in its process as being both strong and clean and offering minimum disturbance to the metal, would withstand hardening and tempering. The only possible question is whether, in the light of anything disclosed or not disclosed in the German patent, it was obvious that tempering the blade, after hardening, would remedy the deleterious action of the hardening temperature upon the particular blade metal used in plaintiff's patent. Here plaintiff, conceivably, may have a point.
Plaintiff's brief says, "The German patent relates to a tool with an entirely different material than the 6150 steel of the saw band made by the method of the patent in suit." As usual, there is no evidence that the German steel and plaintiff's steel are "entirely different." On the other hand, there is no testimony to support the conclusion of defendant's counsel that they are substantially the same. Defendant points to the fact that the German soft steel is described as having carbon, manganese and silicon, and that 6150 steel also contains carbon, manganese and silicon. Plaintiff replies that 6150 steel contains a smaller percentage of these materials. Defendant answers that the German patent describes the percentages merely as "particularly" successful examples. Defendant further points out that plaintiff's patent is not limited to steel whose characteristics measure 6150. "Those skilled in the art will understand that steel alloys other than this one will meet the requirements and hence can be employed." The actual claims of the patent do not specify the composition of the blade steel at all, but merely say "a flexible alloy steel band."
While, if we were obliged to decide, we might think that the defendant has well the better of this argument, we will heed our own advice, United Shoe Mach. Corp. v. Industrial Shoe Mach. Corp., supra, and will not decide a question calling for expert knowledge without the benefit of expert opinion. We accordingly vacate the judgment, and the findings of the court below except those expressly confirmed in this opinion, and remand the case for a new trial on the following issues. Taking the German patent as referring to band saw blades, did it indicate to one skilled in the art that the curative effect of the tempering would apply in the case of blade steels, (1) approximating 6150, or (2) reasonably found to be within plaintiff's patent, having in mind the broad language thereof, or (3) did it at least sufficiently suggest such possibility, so that to try its effect on such steels would not be inventive? If any of these questions are answered in the affirmative, judgment is to be entered for the defendant of invalidity of the patent.
Otherwise, judgment for the plaintiff. However, it must be obvious from the foregoing, that no judgment for the plaintiff should foreclose the defendant from employing the process so long as it uses a flexible alloy steel band that falls within the proper scope or intendment of the German patent.
Since the defendant was essentially successful on this appeal, costs to it.