101 F.3d 714
HERR-VOSS CORPORATION, Plaintiff-Appellant,
DELTA BRANDS, INC., and Samuel F. Savariego, Defendants-Appellees.
United States Court of Appeals, Federal Circuit.
Nov. 8, 1996.
NOTICE: Federal Circuit Local Rule 47.6(b) states that opinions and orders which are designated as not citable as precedent shall not be employed or cited as precedent. This does not preclude assertion of issues of claim preclusion, issue preclusion, judicial estoppel, law of the case or the like based on a decision of the Court rendered in a nonprecedential opinion or order.
Before RICH, PLAGER, and CLEVENGER, Circuit Judges.
CLEVENGER, Circuit Judge.
Herr-Voss Corporation (Herr-Voss) appeals a decision of the district court for the Northern District of Texas holding that its U.S. Patent No. 4,614,101 is invalid and not infringed by a slitting line sold by defendant Delta Brands, Inc. (Delta). We affirm-in-part and vacate-in-part.
* Strip hot- and cold-rolled steel is manufactured in very long lengths and in widths up to 12 feet. Because of its long length, strip-rolled steel is usually stored in coils. In addition, customer preference frequently dictates that these wide coils be slit into several narrower coils. Slitting lines employed for this purpose generally comprise a rotating payoff reel (or uncoiler) which supports the wide coil and feeds the strip to a slitter. The slitter employs pairs of opposed circular knives to slit the strip lengthwise into multiple strands of desired widths. The slit strands are then rewound on a recoiler (or mandrel) into narrower coils of strip-rolled steel.
Wide strip is not of uniform thickness across its width. Instead, the strip is generally thickest near the middle and tapers off towards the strip's edges. This effect is called "crowning." Crowning makes recoiling of the slit strands difficult. As a result of crowning, the strands formed from the middle portions of the wide strip are, generally speaking, thicker than those formed from the wide strip's edges. Because of this differential in thickness, the thicker inner strands build up coils of greater diameter on the recoiler than the coils formed by the outer thinner strands. These thinner outer coils are unable to maintain as high a tension in the outer strands as the thicker inner coils do in the inner strands. This causes the outer strands to droop and ultimately causes the outer coils to be wound less tightly than the inner coils.
Augustine A. Fornataro, president of Herr-Voss, solved this problem by recognizing that the problem itself (i.e., the differential in tension between the inner and outer strands) could provide the means to a solution. Fornataro recognized that one might exploit the greater tension in the inner strands to differentially stretch and, thus, thin the inner strands more than the outer strands. This results in the inner and outer strands being more nearly of uniform thickness, thus permitting the inner and outer coils to rewind more evenly. Fornataro further recognized, however, that the tension provided by the recoiler was not, by itself, sufficient to elongate the slit strands. Therefore, Fornataro additionally fed each of the strands through a device called a "strand tensioner." The strand tensioner comprises two layers of parallel work rolls, an upper layer and a lower layer, which alternately bend the slit strands upwardly and downwardly in a process known as "reverse bending." As Fornataro later explained in his specification:
While tension [from the recoiling] alone below the tensile yield strength of the metal will not elongate and thin the strands, tension in combination with repeated bending reversals of the strands can elongate them and reduce their thickness. All the strands pass between the same bending rolls, but the bending strain reversals produce a greater tension increase in the thicker strands than on the thinner strands. Furthermore, any increase in the diameter of the coils of thicker material will tend to impose additional tension on the thicker strands. The net effect is to elongate and thin the thicker strands relative to the thin, which is an object of this invention.
For his invention, Fornataro was awarded U.S. Patent No. 4,614,101 (the '101 patent). Claim 1, the only independent claim of the patent, recites:
1. The method of coiling into uniformly tight coils multiple metal strands slit from strip varying in thickness from edge to center comprising working the multiple strands selectively by reverse bending between the same rolls, so as to reduce and elongate the thicker strands more than the thinner strands and then coiling the strands in multiple at the same rotational speed.
On May 4, 1992, Herr-Voss, assignee of Fornataro, sued Delta for inducing infringement of its '101 patent, asserting that Delta had sold a slitting line designed to perform the claimed method. Delta defended the action by arguing that, under a proper construction of the claims, its slitting line did not perform the claimed method, and, in the alternative, that the '101 patent was invalid for failure to disclose the best mode, for obviousness, and under the on sale bar of 35 U.S.C. § 102(b). After a six-day bench trial, the district court agreed with Delta on all grounds and entered judgment in its favor. This appeal followed. We have jurisdiction to hear this appeal under 28 U.S.C. § 1295(a)(1) (1994).
A determination of literal infringement requires a two step analysis. The first step is to construe the claim. The second is to compare the properly construed claim to the allegedly infringing process. Markman v. Westview Instruments, Inc., 52 F.3d 967, 976, 34 USPQ2d 1321, 1326 (Fed.Cir.1995) (in banc), aff'd, 116 S.Ct. 1384 (1996). The first step, claim construction, is a matter of law which we review de novo. Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576, 1582, 39 USPQ2d 1573, 1576 (Fed.Cir.1996). Infringement vel non is reviewed for clear error. Amhil Enters. Ltd. v. Wawa, Inc., 81 F.3d 1554, 1562, 38 USPQ2d 1471, 1476 (Fed.Cir.1996).
The district court concluded that Delta's slitting line did not infringe Herr-Voss's patented method on two grounds. The first ground rested on the district court's interpretation of a disputed term in the claim. The claim recites the step of:
working the multiple strands selectively by reverse bending between the same rolls, so as to reduce and elongate the thicker strands more than the thinner strands....
(Emphasis added.) At trial, the parties vigorously disputed whether the elongation recited in the claim referred to elastic (i.e., temporary) or plastic (i.e., permanent) elongation. Under the former interpretation, Delta's slitting line does not infringe the claimed method since, as both parties concede, Delta's line effects plastic elongation of the strands.
At the conclusion of the trial the district court adopted Delta's proposed claim interpretation and held that the claimed method requires elastic elongation of the slit strands. The district court reached this conclusion primarily on the basis of a piece of prosecution history in which Herr-Voss distinguished its invention from an earlier patent issued to Kinnavy. As Herr-Voss explained to the patent office:
The method of selectively working the strands comprises subjecting all strands to reverse bending while applying tension thereto less than the tensile yield strength of the strands. The thicker strands are thereby elongated more than the thinner strands.
The Kinnavy patent discloses a method of permanently elongating strip by applying tension thereto and reverse bending it so as to exceed the yield strength of the metal. It does not deal with the tensioning of slit strip in multiple.
Generally speaking, applying tension to metal below the tensile yield strength results in only elastic deformation. The district court therefore inferred from the above representation that Herr-Voss's invention differed from Kinnavy in that Kinnavy concerned plastic deformation while the invention concerned elastic deformation.
Having reviewed the '101 patent and those portions of the prosecution history and trial transcript made part of the appellate record, and especially the testimony of the inventor, Fornataro, we are convinced that the district court erred in its claim interpretation. We hold that the term "elongate" in the claim refers to plastic, rather than elastic, deformation.
Most persuasive on this point is the sentence in the specification which states:
While tension [from the recoiling] alone below the tensile yield strength of the metal will not elongate and thin the strands, tension in combination with repeated bending reversals of the strands can elongate them and reduce their thickness.
As Delta's expert witness, Dr. Watson, conceded at trial, this sentence is incorrect if the elongation being described is elastic elongation because tension below the yield strength of the metal will elastically elongate the metal. Thus, the patent specification itself uses the word "elongate" to refer to plastic elongation.
The aforementioned representation made by Herr-Voss during prosecution is not inconsistent with our interpretation of "elongate." That representation merely states that in Kinnavy the combination of tension and reverse bending exceeds the yield strength of the metal, while in Herr-Voss's invention the tension by itself is less than the metal's yield strength. Consequently, it does not support the district court's holding that the word "elongate" refers to elastic elongation. Instead, as the district court also recognized, this representation was made to distinguish Kinnavy on the basis that the Kinnavy reference worked only with a single strip of uncut metal whereas the '101 patent applies to multiple strands of slit metal. We therefore cannot affirm the district court's holding of noninfringement on the first ground articulated by the district court.
In its second ground, the district court focused on the language in the claim requiring that the reverse bending of the strands contribute to their elongation ("reverse bending ... so as to ... elongate the thicker strands"). This limitation, the district found, was absent from the method performed by Delta's accused slitting line. As the district court explained:
The '101 patent effectuates the elongation and thinning of metal strands by combining tensions from the pay-off reel, slitter, and recoiler with the tension created in the reverse bending. The accused product cannot be found to read on to the '101 patent because the accused process achieves thinning and elongation through the application of sheer tension above the yield strength in the pull of the recoiler.
In its brief, Herr-Voss argues that the record contains no evidence to support the district court's finding that Delta's slitting line stretches the thicker strands using "sheer tension ... in the pull of the recoiler" rather than a combination of that tension with the reverse bending as claimed in the '101 patent. In fact, Herr-Voss adds, what evidence there is supports the opposite conclusion. Specifically, it points to testimony by Mr. Savariego, owner of Delta, in which he specifically admits that his slitting line employs reverse bending in order to decrease the amount of tension necessary to deform the steel. In response, Delta points to no evidence to support the district court's factual finding that reverse bending does not contribute to elongation of the thicker strands in Delta's slitting line.1 Consequently, we must conclude that this finding by the district court is unsupported by the evidence and is thus clearly erroneous. Since we find that both grounds of noninfringement relied on by the district court were erroneous, we vacate the noninfringement portion of the district court's judgment.
The district court concluded that the patent was invalid for failure to disclose the best mode because the specification does not disclose two types of adjustments to the strand tensioner.2 The two types of adjustments are known as "independent" adjustments3 and "entry and exit" adjustments.4 Strand tensioners sold commercially by Herr-Voss possess the two sets of adjustments.
A determination that a patent is invalid for failure to disclose the best mode necessarily rests on two underlying factual findings which we review for clear error. Engel Indus., Inc. v. Lockformer Co., 946 F.2d 1528, 1531, 20 USPQ2d 1300, 1302 (Fed.Cir.1991). First, the factfinder must find that the inventor contemplated a best mode of practicing the invention at the time the patent application was filed. Second, the factfinder must find that the specification does not adequately disclose the best mode in a manner sufficient to permit one of ordinary skill to practice it. United States Gypsum Co. v. National Gypsum Co., 74 F.3d 1209, 1212, 37 USPQ2d 1388, 1390 (Fed.Cir.1996). Here, Herr-Voss's strongest arguments attack the district court's fact findings with respect to the second aspect of the best mode determination.
In particular, although Herr-Voss concedes that the patent does not describe either type of adjustment, it argues that disclosure of these adjustments was not necessary to satisfy the best mode requirement because they constitute routine details which were well known to those skilled in the art. As we recently explained in Great Northern Corp. v. Henry Molded Prods., Inc., 94 F.3d 1569, 1572-73, 39 USPQ2d 1997, ---- (Fed.Cir.1996), because the level of disclosure required by the statute is defined with reference to those skilled in the art, the inventor need not disclose routine production details, even when those production details constitute the best mode contemplated by the inventor.
Focusing first on the "independent" adjustments, Herr-Voss contends that the evidence uniformly establishes that these adjustments were well known in the art. Consequently, Herr-Voss argues, these adjustments constituted a routine production detail which it was not necessary to disclose. In response, Delta points us to no testimony tending to show that these adjustments were less than routine. We therefore do not affirm that portion of the district court's best mode determination based on Herr-Voss's failure to disclose the independent adjustments.
Herr-Voss similarly argues that the "exit and entry" adjustments were well known in the art and thus constitute a routine production detail not requiring disclosure. Here, however, the evidence of record is less one-sided. In its opinion, the district court noted a test report prepared by the inventor Fornataro during testing of an early prototype designed to perform the patented method. In that test report, Fornataro stated that "[t]he exit opening of the tensioner and the entry opening must be readily adjustable." At trial, Delta elicited an opinion from its expert witness, Dr. Watson, regarding the significance of finding the above-quoted sentence in a test report:
Q. Now, there's been a contention made in this case that Mr. Fornataro didn't have to disclose that the entry and exit openings had to be adjustable because everybody knew that already. How would the way this report is written affect a conclusion like that?
A. If it is common knowledge and everyone knows it I don't think there is even any call to include it in a conclusion in an important document like a test report.
Herr-Voss had an opportunity to cross-examine Dr. Watson but did not utilize that opportunity to attack this aspect of his testimony. Instead, Herr-Voss points to testimony by two of its witnesses that standard "roller levelers" have been supplied with entry and exit adjustments for as long as 70 years and that these adjustments are extremely well known in the art. In and of itself, however, this testimony does little to promote Herr-Voss's position.
It may well be that standard roller levelers have long been supplied with entry and exit adjustments. It may also be, as Herr-Voss states in its brief, that Herr-Voss utilizes a slightly modified5 off-the-shelf roller leveler as the strand tensioner in its slitting lines. But the patent says nothing about using a standard roller leveler as the strand tensioner. It merely teaches constructing a strand tensioner from a plurality of rolls. It does not necessarily follow from the fact that standard roller levelers are typically provided with these adjustments that one of ordinary skill constructing the invention from the specification would know to provide the strand tensioner therein described with entry and exit adjustments. As the test report shows, the inventor himself did not recognize the need for these adjustments until he tested the invention.
Given Dr. Watson's testimony, we cannot say that the district court clearly erred in finding that the entry and exit adjustments were not routine. Consequently, we affirm the district court's determination that failure to disclose these adjustments renders the patent invalid for failure to disclose the best mode.
We have also considered Herr-Voss's other challenges to the district court's best mode determination, but find them without merit. To begin with, Herr-Voss quibbles with the level of specificity in the district court's findings of fact. In particular, Herr-Voss complains that the district court never explicitly found that Fornataro contemplated the adjustments as the best mode on the day he filed his patent application. In addition, Herr-Voss criticizes the district court's failure to make an explicit finding as to whether the two types of adjustments were well known in the art.
It is true that Rule 52 of the Federal Rules of Civil Procedure requires the district court to support its judgment with findings of fact. Rule 52, however, does not require encyclopedic recitation of every underlying fact relied on by the district court in reaching its decision. All that is required is a level of specificity sufficient to enable a reviewing court to understand the district court's reasoning and to ensure the integrity of the factfinding function. See Oakley, Inc. v. International Tropic-Cal, Inc., 923 F.2d 167, 168, 17 USPQ2d 1401, 1403 (Fed.Cir.1991). As one commentator has put it: "In enforcing the requirement, the courts generally take a pragmatic approach, asking not whether the findings are perfect, detailed, pretty, or decisive, as long as adequate review is possible." 1 Steven Alan Childress & Martha S. Davis, Federal Standards of Review § 2.11, at 2-80 (2d ed. 1992). Here, while we agree with Herr-Voss that the district court's findings of fact could have been more clearly articulated, we believe that the level of specificity in the findings is sufficient to ensure adequate review.
In particular, although the district court did not specifically find that Fornataro contemplated the adjustments as the best mode on the day he filed his patent application, the district court's findings as a whole reasonably emit this inference since the district court found that Fornataro knew that these adjustments were necessary prior to his filing date.
Similarly, although the district court did not specifically find that the entry and exit adjustments were not routine, it is reasonable to draw this inference from the findings as a whole. In paragraph 43 of its findings, the district court noted that "[Herr-Voss argues] that the stated adjustments were old and well known in the art, thus, there was no requirement that they be disclosed in the '101 patent." Then, in paragraph 44, the district court referred to the test report prepared by Fornataro in 1977. As noted above, Delta relied on this test report at trial to establish that the adjustments were not routine. From the juxtaposition of these two paragraphs, and the district court's ultimate conclusion that the '101 patent was invalid for failure to disclose the best mode, we believe it reasonable to infer that the district court concluded that the adjustments in question were not routine.
Finally, Herr-Voss argues that although the two sets of adjustments might constitute the best mode had the '101 patent claimed a strand tensioning apparatus, these adjustments cannot constitute the best mode of its claimed method. We disagree. Title 35, section 112 requires the inventor to disclose the best mode of practicing the invention, whether claimed as an apparatus or as a method. The '101 patent describes a method of creating uniformly tight coils of slit strip. One of the steps in that method consists of subjecting the slit strands to reverse bending. If there was a preferred mode of performing this step, i.e., performing the reverse bending on an apparatus provided with entry and exit adjustments, section 112 required its disclosure. Therefore, we affirm the district court's determination that the '101 patent was invalid for failure to disclose the best mode of practicing the invention.
For the reasons described above, the district court's judgment is affirmed-in-part and vacated-in-part.
At oral argument, counsel for Delta sought to support the district court's decision on this point by arguing that the claim requires contributions of "tension" from both the reverse bending and the recoiler, while in Delta's slitting line, the reverse bending serves only to reduce the amount of tension required to permanently elongate the strands. In our opinion, this argument is more semantic than substantive. Whether one calls the contribution from the reverse bending "tension" or whether one views the contribution from the reverse bending as decreasing the amount of tension necessary to achieve permanent elongation, the method steps performed in practicing the invention are the same. Thus, in both the claimed invention and Delta's slitting line, a combination of reverse bending and tension from the recoiler is necessary to permanently elongate the strands
Because we affirm the district court's decision on this ground, we do not reach the question of whether the claims are otherwise invalid, as the district court held
Work rolls like the ones which form the strand tensioner in Herr-Voss's slitting line are frequently provided with backup rolls to provide the work rolls with structural support. Indeed, the '101 patent itself notes: "Rolls 15 and 16 may be provided with back-up rolls to prevent roll bending." These back up rolls may be provided with what are called "independent adjustments." These adjustments can be used to bend the work rolls in selected zones in order to vary the pressure exerted by the work rolls on the strip
Although not entirely clear from the record, it appears that these adjustments permit the first (i.e., entry) and last (i.e., exit) rolls of the strand tensioner to be independently adjusted
Standard roller levelers function much like the strand tensioner described in the '101 patent but are generally self-driven by a motor. When used as a strand tensioner, the roller leveler's motor is removed and the strands are pulled through the roller leveler by the recoiler