Opinion
Civ. A. No. 74-264.
March 3, 1980.
James M. Mulligan, Jr., Connolly, Bove Lodge, Wilmington, Del., Walter D. Ames, and Robert J. Lasker, Watson, Cole, Grindle Watson, Washington, D.C., for plaintiff.
Charles S. Crompton, Jr., Potter, Anderson Corroon, Wilmington, Del., John M. Calimafde, and Marvin N. Gordon, Hopgood, Calimafde, Kalil, Blaustein Lieberman, New York City, for defendant.
OPINION
The patent-in-suit, United States Patent No. 3,799,532 ("the '532 patent"), claims a connector used to join the ends of coil springs forming the media of a filter unit. The holder of this patent is Komline-Sanderson Engineering Corporation ("K-S"). It here asserts that Ingersoll-Rand Company ("I-R") is selling rotary drum vacuum filters utilizing coil spring connectors which infringe the '532 patent.
This Court has subject matter jurisdiction by virtue of 28 U.S.C. § 1338(a). Venue is located in the District of Delaware by reason of I-R's incorporation here.
The '532 patent was issued to K-S on May 2, 1974, upon an application filed in the name of William L. Schlegel on March 16, 1973. The specifications of the patent concede that both the filter and the coil spring filtering elements referred to therein were a part of the prior art at the time of the alleged invention. Indeed, K-S held the basic patents on the coil spring filter until they expired sometime in 1972. Prior to that expiration, K-S possessed a virtual monopoly on this type of filter unit.
In the filter units of both K-S and I-R, a plurality of generally parallel and contiguous, endless springs form a filter belt through which the liquid to be filtered is drawn into the interior of a vacuum drum. The liquid passes through the interstices between the adjoining coils of each spring and a filter cake of solid materials is formed on the outer surface of the spring belt. This cake is discharged during passage of this belt around a discharge roll which together with smaller rollers at the other end of the filter unit defines the circuit traveled by the endless springs. Each spring, after being placed around the rollers in an appropriate location relative to the other springs, must have its ends connected in some manner so that it forms an endless loop.
The desirable characteristics of a connector for use in such a filter are described in the following segment of the patent-in-suit:
The present invention is an improvement on the connector disclosed in the application of William L. Schlegel, Ser. No. 194,168. It is extremely important in filters of the type above described, that the connectors by means of which the ends of the springs are connected to form endless loops, be capable of permitting uniform flexing of all portions of each spring loop, including those portions at and adjacent the connector, as the spring passes repeatedly around the various rolls and drums defining its endless circuit. Otherwise, if the connection is such as to result in a sharper bending or curving of the spring, adjacent to the connector than occurs elsewhere throughout the length of the spring, this will eventually result in fatigue and breakage, thus forming a void through which unfiltered material might pass to contaminate the resulting filtrate.
Moreover, it is desirable that the connectors be capable of permitting relative rotation between the interconnected ends of the springs, so that any stored torque within them may be removed by untwisting to thus avoid rolling and relative displacement tendencies.
The connector claimed in Claim 1 of the patent-in-suit consists of a pair of "generally cylindrical plugs proportioned for free co-axial reception within the ends of the spring" and a "freely flexible tension means affixed to and connecting the . . . plugs in a fixedly spaced relation for free angular movement by flexing of said tension means and for a limited degree of relative rotary movement." The plugs are constrained from withdrawing from the spring ends by bending the ends of the wire forming the coil spring in an inwardly direction after inserting the plugs so that these ends engage the rear surface of each plug.
The claims of the '532 patent read as follows:
1. A connector for coil spring sections forming the filter media of a filter unit, comprising a pair of generally cylindrical plugs proportioned for free coaxial reception within the ends of the coil spring sections to be interconnected; said plugs respectively having relatively opposed end faces in radial planes thereof for axially abutting engagement by radially inwardly directed portions of the respective spring sections; and freely flexible and twistable tension means affixed to and interconnecting the said plugs in fixedly spaced relation for free relative angular movement by flexing of said tension means and for a limited degree of relative rotary movement.
2. A connector as defined in claim 1, in which said tension means comprises a braided wire cable.
3. A connector as defined in claim 1, in which said plugs are formed with relatively aligned sockets opening axially through the central portions of their respective said end faces, the opposite ends of said tension means being fixedly secured in the respective said sockets.
4. A connector as defined in claim 3, in which said plugs are of deformable material and radially inwardly deformed around and in tight frictional engagement with the said ends of the tension means.
5. A connector as defined in claim 1, including a resiliently deformable externally cylindrical sleeve disposed on said tension means and occupying the space between said end faces.
6. A connector as defined in claim 5, in which said sleeve is axially compressed between said end faces and in flush engagement therewith, while being of an external diameter substantially equal to that of the interconnected springs.
Claim 2 claims a connector as specified in Claim 1 in which the tension means comprises a braided wire cable. Claim 5 claims a connector as defined in Claim 1, including a "deformable" sleeve disposed on said tension means and occupying the space between the two plugs.
K-S has never offered a connector like that claimed in the '532 patent for sale. Its filter units utilize a "pin" connector which is described hereafter in the discussion of the prior art. I-R, however, is currently selling filter units which employ a connector which consist of cylindrical plugs connected by a short piece of twisted cable. I-R asserts that its twisted cable, because of its relatively short free length, is not "a freely flexible and twistable tension means" and that sufficient flexibility to permit the spring to conform to the curved roller surface is achieved by permitting play between the end plugs and the ends of the spring.
THE PRIOR ART
During the early years of its monopoly on the spring filter, K-S connected the ends of the springs with rigid connectors; first with a threaded, screw-type connector and, somewhat later, with a "dumbbell shaped connector". The dumbbell connector had two metal end plugs separated by a narrow diameter, rigid shank. The plugs of this connector were generally cylindrical in shape and tapered at the ends like the plugs of the connector disclosed in the patent-in-suit. The dumbbell connector was retained in the spring by crimping the wire end of the springs inward after insertion of the plugs so that these ends engaged the rear surface of each plug.
This connector was admitted into evidence as PX-8. U.S. Patent No. 3,482,462 issued to Dahlem December 9, 1969 discloses a somewhat similar dumbbell type connector.
Following the dumbbell connector, K-S, through William Schlegel, the inventor of the '532 patent, developed what was referred to at trial as a "pin-type", or "articulated", connector. This connector, which K-S introduced in 1971, is the subject of U.S. Patent No. 3,701,432 ("the '432 patent"). PX-10. It consists of two metal plugs connected by a narrow diameter, rigid shank or pin. The spring ends are secured to this pin-type connector in the same manner as in the case of the dumbbell connector. It differs from the earlier connector, however, in that the shank is inserted through holes in the end plugs which have a larger diameter than that of the shank so that the plugs move freely about the shank. This movement permits the plugs to rotate around the longitudinal axis of the pin and to alter their position relative to one another so that the connector can accommodate the curved surface of the rollers without imposing stresses on the spring. The specifications of the '432 patent contain the following relevant teachings:
. . . it has been observed by the undersigned inventor that where the adjoining spring ends are connected by conventional rigid plug type connectors such as heretofore employed, any breakage or failure of a spring which occurs, will invariably be closely adjacent to one of the said connectors. It is believed that such failure arises from the fact that when the spring attempts to conform to the curvature of a roller in passing therearound, the rigid connector plug interconnecting adjacent portions of the spring is unable to assume a corresponding curvature. Therefore adjacent the ends of the rigid plug, the wire filament from which the spring is formed, is subjected to a relatively higher degree of bending or twisting than occurs at other locations along the length of the spring. The repeated flexing at the same location is believed to result in a work hardening and crystalizing action which renders the stainless steel spring filament subject to corrosion and which eventually leads to breakage of the spring at the particular location.
With the foregoing in mind, the present invention comprises an articulated connector or coupling which, when used to interconnect the ends of the coil springs or spring sections incorporated in the filter media, enables the coupled end portions of the springs to conform more closely than heretofore to the curvature of a roll about which the spring is trained, while minimizing any localized bending or twisting of the spring filament as the spring is flexed in passing on to and moving from each roll and/or the drum. . . .
Schlegel was not the first, however, to appreciate the advantage of a connector for a coil spring which would be sufficiently flexible to permit the spring to conform to the surface of the rollers over which it passed. U.S. Patent No. 1,378,507, issued to Wiegand July 19, 1919, claims a threaded connector for a helical spring belt made of rawhide. The specifications observe:
Belts of this character are very advantageous for newspaper conveyers and other mechanisms, but due to the lack of suitable end connections therefor, the same have heretofore proven troublesome in use. In newspaper conveyers, for example, the belts are passed over sheaves of relatively small diameter and the connections have been subjected to breakage due to the strains and wear thereon in passing over such sheaves.
* * * * *
The connecting means for the extremities of spring 1 comprises a threaded member 3 (Fig. 3) having its opposite ends screwed into said spring extremities. The member 3 may be formed of any suitable material although it is preferably formed of rawhide which is strong and durable and also flexible. . . .
Similarly, U.S. Patent No. 3,157,056, issued to Gray November 17, 1964, claims an "elastic plug type connector" with a threaded surface. The specifications teach as follows:
. . . Experience over many years has proven that . . . [the present] method of splicing the ends together, while workable, leaves much to be desired for the simple reason that belt failures almost always take place at the splice. It seems that a splice of this type is considerably stiffer than the remainder of the belt and the continued flexing at the splice which takes place as it passes around the rollers causes premature failure.
It has now been found in accordance with the teaching of the present invention that this and other difficulties can be substantially eliminated by . . . splicing [the] . . . loose-wound end sections together with a flexible threaded connector having helical grooves on the surface thereof adapted to screw into the loose-wound coils. The resulting endless spring belt has several distinct advantages in that the splice has approximately the same degree of flexibility as the close-wound helix that constitutes the major portion of the belt. . . .
The teachings of the prior art discussed thus far pertain to connectors utilized to form endless helical springs which together comprise a belt. The pertinent art is not this limited in scope, however. Given the nature of the problem to which the patent-in-suit is addressed, the scope of the pertinent art is at least broad enough to encompass techniques employed to connect endless belts of other kinds which are required to travel around rollers. Thus the relevant prior art includes U.S. Patent No. 2,995,045, issued to Marty August 8, 1961, which claims a connector for joining the ends of flexible, power transmitting belts, particularly V-belts. This connector consists of generally cylindrical plugs which are proportioned for reception with holes in the ends of the belt and which are connected by a flexible tension means. Specifically, Marty suggests a tension means with "a steel braided wire core" which will result in the connector being "substantially as flexible laterally as the belt itself so as to have little or no effect on the flexibility of the completed belt."
Geo. J. Meyer Manufacturing Co. v. San Marino Electronic Corp., 422 F.2d 1285, 1288 (9th Cir. 1970). ". . . the word `art' includes not only the knowledge accumulated with respect to a problem in a particular industry but that accumulated in those scientific fields the techniques of which have been commonly employed to solve problems of a similar kind in the particular and closely related fields." See also In re Shapleigh, 248 F.2d 96, 45 CCPA 705 (1957).
THE DIFFERENCE BETWEEN THE PATENT-IN-SUIT AND THE RELEVANT PRIOR ART
K-S's dumbbell connector is virtually identical in shape to the '532 connector and attaches to the spring ends in exactly the same manner. The only differences are that the dumbbell connector consists of a single element and has a rigid tension means, while the '532 connector consists of three elements and has a "freely flexible and twistable tension" means.
The plugs of the '432 connector are identical to those of the '532 connector and both sets of plugs attach to the spring ends in the same manner. While the designs of these connectors recognize and address the same problem and both provide flexibility and twistability as the solution, the '432 connector achieves this solution without a flexible tension means.
The Wiegand and Gray connectors have structures which differ materially from the '532 connector and utilize a distinct principle in securing the spring ends. They do, however, teach the desirability of a flexible and twistable tension means and accomplish this objective through the choice of the material comprising the tension means.
Finally, Marty is very similar in structure to the '532 connector, in that it has generally cylindrical plugs connected by a flexible tension means. It does, however, utilize a different approach to joining the connector with the ends of the belt, relying on friction rather than crimping to prevent withdrawal of the plugs.
LEVEL OF SKILL IN THE ART
While relevant evidence is rather sparse, the record does shed some light on the level of skill in the art. Mr. Schlegel's background and that of Mr. Spencer, who designed the accused connector, suggest that artisans who would be working with the problem addressed by the '532 patent would have college degrees and would possess the skills necessary to design industrial machinery. Mr. Schlegel has a B.S. degree and Mr. Kendall Spencer has a degree in mechanical engineering. Each was employed to do, in Schlegel's words, "general engineering", and had responsibility for the design not only of the connectors but of all of the components of the filter unit.
"SECONDARY" EVIDENCE
I find no "secondary" evidence which is particularly helpful in answering the obviousness question. The only area of application of spring connectors which is touched upon at all in the evidence is the application to helical coil spring filter units. In this area, we know that K-S had a patent monopoly and it is reasonable to expect that little development work occurred by others during a substantial period prior to the date of the alleged invention. There is no evidence that anyone at K-S other than Schlegel worked on perfecting the rigid connector during this period and the record is sketchy concerning the intensity and duration of Schlegel's efforts to solve the problem. We do know that Schlegel had previously found a satisfactory solution to the flexibility problem when he devised the '432 connector.
THE REISSUE APPLICATION
On March 30, 1977, K-S caused an application for Reissue Letters Patent to be filed based on the '532 patent. This reissue application was unchanged from the specifications, drawing and claims of the '532 patent. In a letter accompanying the reissue application each of the patents designated as prior art by I-R was brought to the attention of the Patent and Trademark Office and commented on by applicant. These included Dahlem (claiming a connector similar to the dumbbell connector), the '432 patent, Wiegand, Gray and Marty.
In addition to this prior art, the letter informed the Patent and Trademark Office of the following additional art: "Prior use and sale has been made of connectors of the type disclosed in Schlegel Patent No. 3,701,432 and of the rigid `dumb-bell' type more than one year prior to the filing of the application for the patent-in-suit."
In two office actions, the Patent Office Examiner rejected and finally rejected claims 1 to 6 as unpatentable over Marty in view of Dahlem or Bower under 35 U.S.C. § 103. Applicant appealed. On May 17, 1978, the Patent and Trademark Office Board of Appeals reversed the decision of the Examiner and held that claims 1 through 6 were patentable over the art cited by the Examiner and further stated, "In the formulation of our decision, we have carefully considered the entire record in this appeal, including all of the references which have been made of record during the prosecution of this application."
OBVIOUSNESS
The patent-in-suit is entitled to a presumption of validity and must stand unless its invalidity is demonstrated by clear and convincing proof. Tokyo Shibaura Electric Co. v. Zenith Radio Corp., 548 F.2d 88, 93 (3d Cir. 1977). The action of the PTO Board of Appeals cannot be the final word on validity, however, as our Court of Appeals made clear in Alco Kar Kurb, Inc. v. Ager, 286 F.2d 931 (3d Cir. 1961):
We have not ignored appellant's reminder that a significant presumption of validity attends a patent issued, as this one was, only after the Patent Office Board of Appeals, in a contested proceeding, has considered the nature of the structure and decided that it is a patentable invention [citing cases]. But respect for such considered administrative decision does not justify abdication of the normal judicial function. It still remains the responsibility of the United States courts to see that patent monopoly is enjoyed only pursuant to a correct standard of invention properly applied.
I am persuaded that the '532 patent claims a monopoly which Congress did not intend to exist. While I reach this conclusion reluctantly because it conflicts with that of the Board of Appeals, the evidence before me demonstrates in a clear and convincing manner that an artisan skilled in the relevant art would find the alleged invention obvious from the prior art. The reasoning of the Board simply does not seem to me to speak to the problem presented by the case.
As K-S's expert acknowledged, the dumbbell connector is identical to the '532 connector in all material respects except (1) the former is composed of a single element, while the latter consists of three elements — the two plugs and the tension means; and (2) the former has a rigid tension means and the latter a "flexible and twistable" tension means. (B-45-53). It is not claimed that there is anything other than old technology involved in swagging plugs on a tension means. The question for decision, accordingly, is simply whether artisans skilled in doing general engineering would find it obvious to substitute a flexible and twistable tension means or specifically a braided steel cable, for a rigid tension means in designing a connector for springs intended to travel repetitively around drums, rollers or other round objects. There can be only one answer. Numerous pieces of prior art taught the desirability in such an environment of a connector with sufficient flexibility to permit the connecting portion of the endless springs to conform to the drum or roller surface. Moreover, the prior art, through Marty, specifically taught the use of wire cable as a tension means between two plugs in order to accomplish that flexibility. In addition, we know that a skilled artisan, upon being advised of the desirability of flexibility and being aware of the need for strength, would undoubtedly have considered wire cable as the material likely to solve the problem, since it is well known for these characteristics. (B-161-63). Finally, the evidence shows nothing in the prior art which would discourage such an artisan from making this obvious choice of material. (B-176).
The obviousness of the choice of a "flexible and twistable tension" means in general and cable tension means in particular is demonstrated by the experience at I-R. As earlier indicated, when I-R decided to design a filter unit to market at the expiration of the K-S patent, the design work was assigned to Kendall Spencer. He designed the I-R connector concededly without any knowledge of the '532 connector. He testified that when he was first assigned the project, he "realized the need for a connector that would be strong, would make the lowest discontinuity in the spring in terms of operation, . . . would be long lived, and . . . inexpensive." (C-146). He was aware at the time of the dumbbell connector and the pin type connector, realized that the latter was preferable to the former because of its "inherent flexibility", but was concerned that fine particles of sludge might become lodged in the pin connector over time and render it the equivalent of a rigid connector. (C-153). He had been at work on designing all parts of the filter unit for about four months when he saw a length of aircraft cable in his engineering laboratory and realized it was "exactly the tension member that would do the job." The thought occurred to him early in the afternoon and he completed the design of the I-R connector before 4:30 P.M. that day. (C-151-52). The design seemed "so obvious" to him at the time that he did not even consider suggesting an application for a patent. (C-175).
This case is much like Dale Electronics, Inc. v. R.C.L. Electronics, Inc., 488 F.2d 382 (1st Cir. 1973), where the patent claimed a design for an electrical resistor with a core made of a ceramic known as berylium oxide ("BeO"). The resistor design was known in the prior art. Likewise, BeO was known to be a good electrical insulator and a good thermal conductor, characteristics which were known to be desirable in a resistor core. Not surprisingly, the Court held that the choice of BeO as a material for the resistor core was "obvious to . . . [the patentee] and would have been to anyone skilled in the art." Here the prior art taught that connectors for helical coil springs should be strong and flexible, wire cable was known for strength and flexibility, and this material choice would have been obvious to anyone sufficiently qualified to be working in the field.
This case also resembles another case in which I-R was a litigant. Hughes Tool Co. v. Ingersoll-Rand Co., 437 F.2d 1106 (5th Cir. 1971) (use of tungsten carbide inserts implanted in a drilling bit as an improvement over a single element bit of less durable material.)
The PTO Board of Appeals addressed itself to the question of whether the claims of the '532 patent should be "rejected under 35 U.S.C. § 103 as being unpatentable over Marty in view of either one of Dahlem or Bower." Dahlem is similar to the dumbbell connector; the record in this case does not disclose Bower's connector. Having posed the question in this manner, the Board answered it in the negative, reasoning that to modify Marty by substituting plugs which would attach to springs in the manner suggested by the '532 patent would distort and disable the Marty connector in its intended environment and, accordingly, would not have been obvious to one skilled in the art. The Board observed:
. . . It is our opinion that it would not have been obvious to combine the references as suggested by the examiner in that there is no need in the primary reference for the modification proposed. In Marty the connecting element . . after being inserted in the V-belt, is precluded from removal by the frictional contact between the connector structure and the internal surface of the aperture of the V-belt. There is, therefore, no reason whatsoever for modifying the Marty reference in view of the Dahlem or Bower patents. Further, the examiner's suggested substitution of the Dahlem or Bower spring belt for the V-belt of Marty would have required a complete reconstruction of the primary reference which, in our opinion, would not have been obvious to one of ordinary skill in the art. . . . There must be a reason apparent at the time the invention was made to the person of ordinary skill in the art for applying the teachings at hand, or the use of the teachings as evidence of obviousness will entail prohibited hindsight.
The Board presumably approached the question of obviousness from this perspective because it was the perspective from which the Examiner had approached the issue. If the Board had viewed the matter from the perspective of Schlegel or Spencer, or anyone else attempting to solve the problem which they addressed, however, I believe it would have reached a different conclusion. Such an artisan was working in a field where the dumbbell connector was recognized as a satisfactory way of attaching the ends of springs but was known to have the disadvantage of being rigid. The question is whether a review of Marty, together with the prior art teaching the desirability of flexibility, would have suggested to such an artisan the substitution of a flexible cable tension means for the rigid tension means of the dumbbell connectors. As indicated, I conclude that it would.
Judgment will be entered for I-R.
The Court having reached the conclusion that the '532 patent is invalid under 28 U.S.C. § 103 and I-R having abandoned its claim of fraud on the patent office, no other issues raised by the parties need be addressed.