Full opinion text
OPINION MURRAY M. SCHWARTZ, Chief Judge. Plaintiff Afros, S.p.A. (“Afros”) has sued Krauss-Maffei Corporation (“KMC”) for infringement of its United States Letters Patent No. 4,332,335 (“’335 Patent”). KMC, a wholly-owned subsidiary of Krauss-Maffei Aktiengesellschaft (“KMAG”), denies any infringement and counterclaims for infringement by Afros of its United States Patents No. 3,706,515 (“ ’515 Patent”) and No. 3,975,128 (“ ’128 Patent”). All three patents relate to high-pressure impingement mixing used to produce polyurethane products. The parties manufacture or distribute mixing heads for use in both open and closed pour molds. The Court conducted a nine-day trial on liability issues that concluded on February 27,1987. By the parties’ consent, the damages issue has been reserved for a later proceeding. Post-trial argument occurred on April 29 after the filing of proposed findings of fact and briefs. After careful consideration of the sufficiency, weight, and credibility of the witnesses’ testimony, their demeanor on the stand, documentary evidence admitted at trial, and the parties’ post-trial submissions, the following constitutes the Court’s factual and legal conclusions in accordance with Federal Rule of Civil Procedure 52(a). Because the parties each allege infringement of its patents, the Court considers those issues separately. The Court holds the ’335 Patent valid and literally infringed by KMC’s mixing heads. Defendant’s infringement is also willful under 35 U.S.C. § 284, and an exceptional case under 35 U.S.C. § 285. On the counterclaim, the Court holds the assignment of the ’515 and ’128 Patent rights between KMAG and KMC ineffective to grant KMC standing to bring the counterclaim. In the alternative, the Court holds the two patents are valid but not infringed. 1. BACKGROUND Oversimplified, flexible foam polyurethane produced by the mixing heads of the parties are the result of a process whereby two polymers are mixed together and deposited in a mold in which they harden to produce the desired item with the correct shape and weight. The mixing procedure, called impingement mixing, takes place in a mixing head. Trial Transcript (“Tr.”) 86. The two materials used to form polyurethane, polyol and isocyanate, must be throughly mixed if the product is to be acceptable. Tr. 85. Polymerization begins when the chemicals first interact, but the completion of the process takes varying amounts of time depending on the exact polymers being impinged. The mixing process requires atomization of the components in a narrow area, where a high degree of turbulence is generated. Tr. 59. The components are delivered under high-pressure, usually at about 200 atmospheres or “bars,” Tr. 68, and reach the mixing chamber through ports in the mixing chamber. The mechanism through which the polyol and isocyanate enter the impingement zone is variously called “in-feed ducts,” “transfer ports,” and “inlet nozzles.” For the purposes of this opinion, the Court will use the term inlet nozzles to describe the apparatus for delivering components to the mixing chamber. An external pump maintains the necessary pressure during periods of impingement and recirculation. Tr. 158. KMAG and Afros manufacture mixing heads which are similar in most respects. Both the Afros heads, designated “FPL,” and KMAG’s, designated “UL,” are L-shape. The L-shape design entails a mixing chamber connected at a 90° angle to a discharge or quieting chamber. The polyurethane flows from the discharge chamber into the mold. The chambers are circular, with the length and diameter of the discharge chamber greater than the mixing chamber. PX 151. The key issue in the design of these competing heads is the closeness of the component inlet nozzles to the discharge chamber in an L-shape head. Under the ’335 Patent, the location of those nozzles is to be “close to” or “in close proximity to” the quieting chamber. PX 1. According to Dr. Carlo Fiorentini, inventor of the '335 Patent device, locating the inlet nozzles close to the discharge chamber permits additional mixing in the component flow as it exits the mixing chamber and the residual kinetic energy causes a rebounding effect off the opposite wall of the discharge chamber because of the L-shape configuration of the chambers. Tr. 187-88. The L-shape heads are the product of nearly a decade of research and development in the Reaction Injection Molding (“RIM” industry. In order to achieve good mixing of the polyol and isocyanate, which have an affinity for one another akin to oil and water, the industry moved toward high-pressure mixing heads in the 1970’s. Tr. 1554. Unlike low-pressure heads, high-pressure devices mix the chemicals outside the product mold and then pour the polyurethane directly into it. One problem the industry faced was how to maintain the necessary degree of pressure when the components were not being impinged. The defendant’s ’515 Patent, which has recirculation grooves carved into the sides of the mixing piston, permits the maintenance of high-pressure in the inlet nozzles whether or not the mixing head is in operation. Tr. 1552-53. When the mixing piston is in a closed position, the recirculation grooves direct the components back through the pump system into a storage area. The pump maintains the requisite pressure for impingement mixing until the piston is retracted and the inlet nozzles are unobstructed. The use of recirculation grooves was an important step in the development of the mixing heads at issue in this litigation. See Tr. 1552. The essential principles for RIM have not changed radically since the early 1970’s. According to defendant’s expert, Dr. Christopher Macosko, the key concepts for high-pressure mixing heads are: 1) High velocity recirculation. 2) Controlling the duration of the mixing and pouring of the fluid. 3) A retractable plunger to clean the chamber that both permits recirculation and terminates impingement. 4) Creating sufficient back pressure in the mixing chamber both to increase mixing efficiency and eliminate air bubbles in the fluid. Tr. 1550, 1553, 1555. The central element in the head is the mixing piston, also called a “control slide.” The piston has two positions, and its movement is controlled by means of a hydraulic pump system. In the forward or closed position, the chemicals recirculate under high-pressure through the recirculation grooves in the piston. When the piston is retracted, in which position the head is “open,” it no longer blocks the inlet nozzles and the chemicals discharge into the mixing chamber. The inlet nozzles for the chemicals are positioned opposite each other and, once the head is open, the chemicals will impinge and form a flow that will be deposited into the mold.' At the point of impingement, the highly turbulent flow created in the mixing head permits the polyol and isocyanate to continue to mix after the point of impingement. The length of time in which the head is open, called the “shot,” depends on the volume necessary to form the product. Tr. 85. It can last from a few seconds up to half a minute. After the shot, the mixing piston moves forward to both clean the mixing chamber, by forcing any residual matter into the discharge chamber, and recirculates the chemicals. The discharge chamber, placed at a 90° angle to the mixing chamber, is similarly equipped with a retractable plunger called the cleaning piston. When the shot begins, the cleaning piston will be fully withdrawn, permitting a continuous flow of material out of the mixing chamber into the discharge chamber. Once the mixing piston advances to close the chamber, the cleaning piston will move down to expel any remaining polyurethane. The face of the mixing piston is curved to permit the two pistons to be closed at the same time. The advantage of high-pressure mix heads is more efficacious mixing than in low-pressure systems. Tr. 1550. The degree of mixing achieved in high-pressure heads will vary, however, depending on a number of factors. There are three major problems to be overcome iii constructing a workable high-pressure mixing head: 1) cleaning the mixing and discharge chambers; 2) minimizing what has variously been termed the transient phenomenon or lead-lag problem; 3) creating a good mix while allowing the flow to become laminar, or quieted, at the point of discharge into the mold. Cleaning the Head: Polyol and isocya-nate are highly reactive components that begin their chemical reaction upon impingement to form polyurethane. The length of the polymerization will depend on the particular polyol used. Any residual liquid in the chamber will harden, impeding the movement of the piston and degrading the quality of later shots. The chamber through which chemicals flow must be thoroughly cleaned, otherwise a gluing effect occurs in the head. The face of the mixing piston can serve this function when it is moved into the closed position if it is highly precisioned, leaving no space between the piston face and the chamber walls for polyurethane to accumulate. The cleaning process can also be achieved through sealing grooves on the outer edge of the piston that allow a small amount of polyurethane to form and thereby close off the space between the piston and the wall of the mixing chamber. Tr. 1461; see infra, text at 1434-1435. In L-shape heads, which have two chambers at 90° angles, a second piston is required to clean the discharge chamber. Tr. 183-84, 1651. In addition to thoroughly removing residual polyurethane, the cleaning piston must be moved rapidly in order to minimize the possibility of the compound hardening before it can be cleaned out of the chamber. Transient Phenomenon: The design of the mixing piston requires that there be an area between the end of the longitudinal recirculation groove and the piston’s face that permits neither recirculation nor impingement of the components. As the piston retracts from a closed to open position, the inlet nozzles are completely blocked, and the pressure in the lines increases. Tr. 112. How high the pressure builds depends on the area between the recirculation grooves and the piston face and the speed with which the piston retracts. This period, called the transient time, will only last milliseconds, but the longer the blockage the greater the effect on mixing because the components are delivered at a higher than optimum pressure that can negatively affect the mixing. Moreover, as the piston face slides past the inlet nozzles, the flow of components into the chamber is directed away from the opposed nozzle, thereby lessening the impingement. This misdirection of components will also occur as the piston moves to a closed position. The off-ratio portions in the flow can cause a poorly formed product that exhibits defects in a localized area as a result of the transient phenomena at the beginning and end of the shot. Tr. 1803. Mix/Flow Problems: The development of high-pressure impingement mixing requires the coordination of two competing objectives: creation of a highly turbulent flow in the head to permit maximum mixing of the components, and a laminar flow out of the head to the mold. If the flow is too chaotic at the point of ejection, then the polyurethane fluid will not fill the mold evenly, creating a poorly formed product. If insufficient turbulence occurs in the head, then the components will not be properly mixed and, again, a poorly formed product results. It is the delicate balancing of these two objectives that primarily determines the success of the operation of a mixing head. The diameter of the mixing chamber has an important effect on the degree of mixing that will occur in the mixing' chamber. The greater the distance between the inlet nozzles, the less the components will mix. If the diameter is too small, however, the flow may not be sufficiently quieted at the point of discharge into the mold to properly form the product. Tr. 456-60. Prior to 1978, the predominant design of high-pressure mixing heads was straight-line. This type of head had a single piston to clean the chamber and recirculate the components. The polyurethane fluid is mixed and discharged from the same chamber. Manufacturers generally used the devices in a closed-mold system, with the head directly attached to the mold. To eliminate the transient phenomenon, an aft-ermixer attached to the mold would divert fluid from the beginning and end of the shot into an area away from the actual product mold. The undesirable polyurethane from these segments of the shot contained most of the effects of the transient times. The aftermixer would also serve to further mix the flow as it exited the head by creating an area of blockage that generated additional turbulence. After the polyurethane hardened, the mold was opened and the residue in the after-mixer was discarded. In addition to possible problems resulting from the size of the mixing chamber, the straight-line technology faced two additional drawbacks. First, in closed-mold systems the mixing head could only be used for discrete operations because the mold had to be removed, cleaned, and reattached before another shot could commence. Attempted application of the heads in open pour molds, where the fluid discharged into a free standing mold not attached to the head and with no aftermixer, proved generally unsuccessful because of poor mixing and splashing of the fluid in the mold. Tr. 170-71. The closed-mold method, therefore, increased the time necessary to form a polyurethane product. Second, the residue in the aftermixer was discarded after the mold hardened, an uneconomical use of the polyurethane. Tr. 594. Nevertheless, the straight-line heads were commercially viable, but the industry continued to work on overcoming problems with poor mixing and the need for a head for use with open-pour molds. Afros and KMAG developed and marketed straight-line heads during the 1970s. In 1978, Dr. Fiorentini submitted a patent application for an L-shape mixing head. Two earlier patents, the KMAG ’128 Patent and the Upjohn Company’s United States Patent No. 4,155,299 (“Upjohn ’299 Patent”), had previously disclosed a high-pressure L-shape head, but the ’385 Patent cited only the ’128 patent as prior art. The Afros FPL heads, which conform to the description and drawings of the ’335 Patent, Pretrial Order, Docket Item (“Dkt.”) 127, ¶ 7, have short mixing chambers, less than one mixing chamber diameter. This permits the piston to have a shorter stroke, allowing for a light mixing head because the hydraulic pump system need not generate as much force to operate the pistons. Tr. 180. Moreover, the more rapid movement of the mixing piston limits the transient phenomenon. According to Dr. Fiorentini, the key to his invention is the continued mixing of the components in the discharge chamber, caused by locating the inlet nozzles close to the discharge chamber, creating a short mixing chamber, in combination with the L-shape configuration. Tr. 187-89. This additional turbulence, which occurs as the polyurethane flow exits the mixing chamber and reaches the walls of the discharge chamber, creates a better mix than in heads with long mixing chambers, either straight-line or L-shape. The transverse placement of the discharge duct, with its greater length and volume as compared to the mixing chamber, permits rapid conversion to a laminar flow despite the added turbulence caused by the design. The KMAG UL heads are virtually identical to the FPL heads, except for the location of the component inlet nozzles. KMC began marketing the UL head in the United States in 1982, Tr. 845, while Afros made its first American sale of the FPL head in 1978. PX 32. Both products have been relatively successful in terms of overall sales, especially in light of stagnant demand in the polyurethane molded products market. The advantage of the L-shape over straight-line heads is their ability to be used for open-pour molds, permitting au-tomization of the manufacturing process. Robotic systems may be employed with L-shape heads, which are more economical because they are less labor-intensive. Products of open pour-systems are at least as good as, and in many cases better than those produced by the earlier straight-line heads. Another advantage is less wasted polyurethane with the elimination of the aftermixer. Plaintiffs evidence shows nearly half its sales have included an entire system for open-pour molding rather than just the FPL head. The heads alone cost between $8,000 and $12,000, and sales have expanded as manufacturers replace straight-line heads with L-shape heads. Tr. 592, 594-95. II. THE PATENTS IN SUIT The three patents in this litigation are stages in the development of high-pressure mixing heads from the straight-line method illustrated in KMC’s ’515 Patent to the L-shape design of Afros’ ’335 Patent. The claims in issue are presented below. A. The ’335 Patent The Patent, entitled “Head for Mixing and Ejecting Interacting Liquid Components, for Molding Plastic Articles,” issued on June 1, 1982. The claims in issue are numbers 1, 6, 8, and 9, which state: 1. A head for continuously mixing and ejecting interacting liquid components into a mold for molding articles made of plastic material, comprising in combination: a body defining a mixing chamber for the components; a discharge duct having an outlet; at least a first and a second infeed duct for injecting individual components into said mixing chamber for a predetermined time period; exit port means for establishing fluid communication between said infeed ducts and said discharge duct, said infeed ducts being provided with outlet orifices that open into said mixing chamber close to said exit port means; and piston means for controlling fluid communication between said infeed ducts and said discharge duct outlet so that fluid communication is established therebetween during substantially the entire predetermined time period when said infeed ducts are injecting components into said mixing chamber so that injected components flow continuously out of said mixing chamber through said discharge duct and are thoroughly mixed during their passage through said discharge duct, said mixing chamber forming an angle of substantially 90° with said discharge duct and having a length shorter than and a volume less than that of said discharge duct so that components injected into said mixing chamber mix with each other and overflow into said discharge duct in a swirling flow. * # * * # * 6. A head for continuously mixing and ejecting predetermined variable amounts of fluid materials, comprising a body defining a mixing chamber; a discharge duct having at one end an outlet portion for ejecting a mixture, said discharge duct having an inlet communicating with said mixing chamber and having an axis extending transverse to an axis of said mixing chamber; a cleaning member able to slide in the discharge duct between recessed and extended positions, in said recessed position said duct inlet being in fluid communication with said duct outlet portion and in said extended position said cleaning member blocking fluid communication between said duct inlet and said duct outlet portion; a first double acting cylinder for moving said cleaning member between said recessed and said extended positions; a first and a second duct opening into said mixing chamber for infeeding individual components, said infeeding ducts being positioned in close proximity to said duct inlet so that mixing action initiated in said mixing chamber continues in said discharge duct; a first and a second duct for recycling the individual components from said mixing chamber; and a slide valve movable between a first position in which selected ones of said infeed and recycling ducts are in fluid communication with each other and fluid communication is blocked between said infeeding ducts and said duct inlet, and a second position in which said infeeding ducts are in fluid communication with said duct inlet and fluid communication is blocked between said infeeding and said recycling ducts; a second double acting cylinder for controlling movement of said slide valve between said first and said second positions; and operating link-up means between said double acting cylinders for moving said slide valve from said first second position after said cleaning member has been moved from said extended position towards said recessed position so that said infeeding ducts, mixing chamber, and discharge duct are in fluid communication with each other during substantially the entire period during which individual components are fed into said mixing chamber so that the components overflow from said mixing chamber into said discharge duct, said operating link-up means comprising a switch operatively associated with said first double acting cylinder, an electrovalve actuated by said switch, a supply circuit of said second double acting cylinder being connected to said electrovalve. 8. A head according to claim 6, wherein the cross-sectional area and volume of said mixing chamber are less than the cross-sectional area and volume of said discharge duct. 9. A head according to claim 1 or 6, wherein the length of mixing chamber is very limited compared with the length of the discharge duct. PX l. B. The Krauss-Maffei ’515 Patent The Patent, entitled “Device for Feeding Flowable Material to a Mold Cavity,” issued on December 19, 1972. The relevant claims in issue are numbers 1-3, which state: 1. An apparatus for charging a mold cavity with a mixture of at least two flowable components, comprising housing means forming a mixing chamber opening into said cavity at one end and provided with at least one pair of inlet ports communicating with said chamber and spaced from said end; and a control slide shiftable along said chamber for simultaneously regulating the flow of said components from said inlet ports into said chamber and out of said chamber into said cavity, said control slide having a first position wherein flow of said components from said chamber into said cavity is blocked and a second position in which flow of said components into said chamber and a mixture of the components into said cavity is permitted; wherein said control slide defines with said housing means respective recirculation channels communicating with said inlet ports in said first position of said control slide, said apparatus further comprising respective pumps communicating with said inlet ports for delivering said components thereto and having intake sides connected with the respective recirculation channels in said first position of said slide. 2. The apparatus defined in claim 1 wherein said control slide is received within said chamber. 3. The apparatus defined in claim 2 wherein said control slide is formed at least at its extremity turned toward said end of said mixing chamber with a cross section corresponding to the cross section of said chamber and an end face adapted to lie flush with a wall of said mold cavity in said first position. DX 8. C. The ’128 Patent The Patent, entitled “System for Filling a Mold with Reactive Synthetic Resin Components,” issued on August 17, 1976. The relevant claims in issue are numbers 1-3 and 7, which state: 1. An apparatus for filing a mold, said apparatus comprising: a housing formed with a quieting chamber opening in one direction into said mold and with a mixing chamber opening in another direction transverse to said one direction into said quieting chamber, said quieting chamber having a cross-sectional area equal to that of said mixing chamber and a width greater than that of the inlet opening into it from said mixing chamber; means operatively associated with said mixing chamber for injecting a pair of reactive components into said mixing chamber in opposing and impinging jets to form a mixture therein; means operatively associated with said mixing chamber for displacing said mixture in said other direction into said quieting chamber; a piston in said quieting chamber; and means operatively associated with said piston for reciprocating said piston in said one direction in said quieting chamber for displacing said mixture from said quieting chamber into said mold, said quieting chamber being of uniform cross-section in said one direction and has an inner surface formed with said inlet opening, said mixing chamber being of uniform cross-section in said other direction and terminating at said opening. 2. The apparatus defined in claim 1 wherein said means for displacing said mixture in said other direction includes a mixing piston in said mixing chamber having an end face and displaceable between an advanced position with said end face flush with said surface and a retracted position with said end face spaced in said mixing chamber from said surface, said piston being formed with passages for recirculating said components in said advanced position. 3.The apparatus defined in claim 2 wherein said end face is so formed that in said advanced position of said mixing piston said end face completely fills said opening and is continuous with said inner surface. * He * * * * 7. The apparatus defined in claim 1 wherein said one direction is perpendicular to said other direction. DX 9. In the prosecution of the patent, the Examiner required an amendment to Claim 1 before he would permit the issuance of the patent. The new language limited the claim by requiring that “said quieting chamber having a cross-sectional area equal to that of said mixing chamber and a....” PX 229 at 71. PART ONE: THE AFROS PATENT III. VALIDITY OF THE ’335 PATENT Plaintiff alleges the KMC UL mixing heads literally infringe its Patent. Defendant argues the invalidity of the ’335 Patent as being obvious under 35 U.S.C. § 103 and indefinite under 35 U.S.C. § 112, second paragraph. In addition, defendant asserts the ’335 patent is unenforceable because of inequitable conduct by Afros before the Patent and Trademark Office (“PTO”). The gravamen of the inequitable conduct charge is plaintiffs failure to cite the Upjohn ’299 Patent as prior art. The Court will determine validity before considering whether defendant’s UL heads infringe the ’335 Patent. A. Obviousness Under 35 U.S.C. § 103, the Court determines “whether the claimed invention would have been (not ‘would be’ ...) obvious at the time the invention was made_” Hybritech, Inc. v. Monoclonal Antibodies, Inc., 802 F.2d 1367, 1379 (Fed.Cir.1986), cert. denied, — U.S. -, 107 S.Ct. 1606, 94 L.Ed.2d 792 (1987) (emphasis in original). Obviousness is a question of law. Panduit Corp. v. Dennison Manufacturing Co., 810 F.2d 1561, 1566-68 (Fed.Cir.), cert. denied, — U.S. -, 107 S.Ct. 2187, 95 L.Ed.2d 843 (1987) (opinion on remand). The principle underlying the Court’s analysis is the presumption of validity for all patents. 35 U.S.C. § 282. See Panduit Corp. v. Dennison Manufacturing Co., 774 F.2d 1082, 1096 (Fed.Cir.1985) (“Patents are born valid.”), vacated, 475 U.S. 809, 106 S.Ct. 1578, 89 L.Ed.2d 817 (1986), aff'd on remand, 810 F.2d 1561 (Fed.Cir.1987). The burden of proving patent invalidity is on the challenger, under the clear and convincing evidence standard. Bausch & Lomb, Inc. v. Barnes-Hind/Hydrocurve, Inc., 796 F.2d 443, 446 (Fed.Cir.1986). Evidence of material uncited art in the patent prosecution does not affect the presumption of validity, although it may effectively make it easier for the challenger to carry the burden of proof. See Lindemann Maschinenfabrik GMBH v. American Hoist & Derrick Co., 730 F.2d 1452, 1459 (Fed.Cir.1984). In order to reach its legal conclusion under § 103, the Court must make the fourfold factual findings required by Graham v. John Deere Co., 383 U.S. 1, 17, 86 S.Ct. 684, 694, 15 L.Ed.2d 545 (1966): 1)the scope and content of the prior art; 2) the difference between prior art and the claims at stake; 3) the level of ordinary skill in the art; 4) objective [secondary] evidence of non-obviousness. Akzo N.V. v. United States International Trade Commission, 808 F.2d 1471, 1480 (Fed.Cir.1986), cert. denied, — U.S. -, 107 S.Ct. 2490, 96 L.Ed.2d 382 (1987); Bausch & Lomb, 796 F.2d at 447; Loctite Corp. v. Ultraseal, Ltd., 781 F.2d 861, 872 (Fed.Cir.1985). Only upon conclusion of the John Deere factual analysis may the Court decide the legal issue of obviousness. 1. Scope and Content of the Prior Art The scope of the prior art is that which is “reasonably pertinent to the particular problem with which the inventor was involved.” Lindemann Maschinenfabrik, 730 F.2d at 1460 (quoting Stratoflex, Inc. v. Aeroquip Corp., 713 F.2d 1530, 1535 (Fed.Cir.1983)). See Bausch & Lomb, 796 F.2d at 449. The defendant argues that, in addition to the '515 Patent, United States Patent No. 4,175,874 (“Schneider ’874 Patent”) and the Upjohn ’299 Patent, both of which plaintiff failed to cite to the Examiner, are relevant prior art. The Upjohn ’299 Patent, entitled “Frothing Method and an Apparatus for Carrying Out the Method,” details a high-pressure mixing head adaptable for use in an L-shape configuration. DX 215. Defendant asserts the inclusion of this prior art, in combination with the references cited to the Examiner renders the ’335 Patent obvious. The Upjohn ’299 Patent is clearly pertinent, in the broad sense of the word, because it exhibits a number of features in both the cited prior art and the ’335 Patent. The Patent’s primary teaching is the introduction of a frothing agent through a diametrically opposed chamber that will mix with the previously mixed polyol and isocyanate at the intersection of the two chambers. At this point, the frothing agent or propellant will interact with the polyurethane flow as both descend through the discharge chamber. The design is essentially L-shape, with a second mixing head attached to provide a discharge chamber. The Upjohn ’299 Patent provides for continuous fluid communication between the two chambers when the head is in operation. The chemical components impinge through diametrically opposed inlet nozzles, recirculate through grooves on the side of the mixing piston when the head is closed, and the mixing chamber is cleaned by a retractable piston. The discharge chamber is larger than the mixing head. The Upjohn ’299 Patent need not, however, be operated solely with the L-shape configuration. The Patent summary states, “The second mixing head can also be removed so that use of the first mixing head in accordance with conventional methods becomes possible.” DX 215, col. 2, lines 66-69. The device’s structure entails attaching perpendicularly two straight-line mixing heads. In creating the polyurethane foam, the initial impingement of the components occurs in a separate mixing chamber to prevent interference in that process by the introduction of the propellant. Tr. 754; DX 215, col. 1, lines 87-52. The Schneider '874 Patent, entitled “High Pressure Mixing Head for Multi-Component Plastics, Particularly Polyurethane,” was not cited to the Examiner. The Patent, granted to Mr. Fritz Schneider, presently employed by KMAG, teaches attaching two or three conventional straight-line mixing heads to a central discharge chamber. DX 164. The mixing chambers, positioned at 180° or 120° angles, depending on whether two or three heads are attached, deposit the polyurethane fluid into the discharge chamber in order to create a laminar flow prior to injection into the mold. DX 164, col. 1, lines 46-54. Recirculation grooves, diametrically opposed component inlet nozzles, and a piston to clean the chamber after the shot are also provided. According to Claim 1, the discharge chamber diameter is “significantly greater than the diameter of the mixing chambers....” The Court must consider the uncited pri- or art together with the relevant citations considered by the Examiner, the ’128 and ’515 Patents, in determining the scope of the prior art. The ’128 Patent, reviewed extensively by the Examiner, discloses an L-shape configuration with two straight-line heads attached at a 90° angle. The discharge chamber, as disclosed in Claim 1 of the ’128 Patent, has a “cross-sectional area equal to that of said mixing chamber and a width greater than that of the inlet opening into it from said mixing cham-ber_” DX 9, col. 4, lines 39-42. Claims 4 and 5 of the Patent provide for either circular or rectangular discharge chambers. If the cross-sectional area of circular chambers is equal, then the width of the “inlet opening” from the mixing to the discharge chambers cannot be different, as Claim 1 requires. The content of the ’128 Patent regarding the relative sizes of the two chambers is subject to conflicting interpretations. See infra, text at 1450-1453. The Patent also provides for recirculation grooves, diametrically opposed component inlet nozzles, and pistons to clean the chambers. The operation, as described in the specifications, is for a batch process, where the fluid fills the mixing chamber before the discharge chamber piston retracts to permit the fluid to fill the mold. DX 9, col. 4, lines 1-23. The Patent is not, however, necessarily limited to the batch operation as nothing in the Claims prohibits fluid communication between the mixing and discharge chambers during operation. The ’515 Patent discloses the recirculation grooves designed to deflect the components back into the external pumps when the piston is closed. The patent exhibits a straight-line mixing head, insofar as the claims are phrased in the singular tense concerning the “control slide” (i.e., piston) and “mixing chamber.” In describing the flow of components, Claim 1 states, “... a control slide shiftable along said chamber for simultaneously regulating the flow of said components from said inlet ports into said chamber and out of said chamber into said [mold] cavity_” DX 8, col. 8, lines 48-52. The obvious teaching of the patent, especially when read in light of the specifications and accompanying drawings, is a single chamber for mixing and discharging the polyurethane. The ’515 Patent has diametrically opposed component inlet nozzles, and the “control slide” performs a cleaning function. DX 8, Claim 3. 2. Differences Between the Prior Art and the Patented Invention In comparing the ’335 Patent with the prior art, two principles guide the analysis: first, the invention as a whole must be considered, and not simply its constituent parts as related to previous patents. Vandenberg v. Dairy Equipment Co., 740 F.2d 1560, 1566 (Fed.Cir.1984); see Akzo N.V. v. United States International Trade Commission, 808 F.2d at 1481 (party challenging validity “cannot pick and choose among individual parts of assorted prior art references” to prove obviousness). Jones v. Hardy, 727 F.2d 1524, 1528 (Fed.Cir.1984) (the differences from the prior art are not what make a device patentable). Second, as the Court of Appeals for the Federal Circuit stated, [virtually all inventions are necessarily combinations of old elements. The notion, therefore, that combination claims can be declared invalid merely upon finding similar elements in separate prior patents would necessarily destroy virtually all patents and cannot be the law under the statute, § 103. Panduit Corp., 810 F.2d at 1575 (footnotes omitted). The task of distinguishing a patented device from the prior art is difficult because cases have upheld patents combining elements from the prior art while others have been found obvious when no prior art reference specifically disclosed the claimed invention. Panduit Corp., 774 F.2d at 1094; compare Fromson v. Advance Offset Plate, Inc., 755 F.2d 1549, 1556 (Fed.Cir.1985) (nonobvious) with Orthopedic Equipment v. United States, 702 F.2d 1005, 1013 (Fed.Cir.1983) (obvious). The review of the prior art demonstrates common features among all L-shape mixing heads regardless of the configuration. First, the ’515 Patent’s recirculation grooves have set the pattern for succeeding high-pressure mixing head devices. Second, the length of the “control slide” effectively dictates the location of the inlet nozzles, which are diametrically opposed in order to effectuate the mixing. Third, any chambers in the device for mixing or discharging polyurethane will be cleaned by a piston designed to expel residual material before the next operation. Fourth, three prior art references disclose an L-shape with separate mixing and discharge chambers, although the Schneider ’874 Patent can also accommodate a triaxial configuration. Fifth, the Upjohn ’299 Patent shows fluid communication between the chambers, and the ’128 Patent’s claims do not prohibit continuous mixing and ejecting of polyurethane instead of a batch process. Sixth, the different diameters of the mixing and discharge chambers, the latter given a larger size in order to ease the transition to a laminar flow, have been specifically incorporated into the Schneider ’874 Patent and Upjohn ’299 Patent. The contradictory language concerning chamber sizes in Claim 1 of the '128 Patent does not, however, teach a design where the discharge chamber is larger than the mixing chamber. The teachings of the prior art patents are relevant in determining whether the ’335 Patent is obvious. The ’515 Patent teaches a straight-line mixing head. The Claims, specifications, and drawings exhibit a device with a single chamber for mixing and injecting the polyurethane. The Upjohn ’299 Patent discloses an L-shape head with fluid communication between the mixing and discharge chamber during operation. The patent’s primary teaching, however, is directed to “frothing.” The Up- john ’299 Patent’s Claim 1 covers a “frothing method for mixing a liquid propellant having a low boiling point with further components for 'production of foamed ma-terial_” DX 215, col. 4, lines 60-62 (emphasis added). The Patent uses the perpendicular discharge chamber to introduce the propellant at a point after the components have been mixed. It is not designed to effect additional mixing except by way of the propellant. The Upjohn ’299 Patent, therefore, does not teach the L-shape as a means for better mixing or elimination of the transient phenomenon by use of a lighter piston with a short stroke. The Schneider ’874 Patent teaches attaching two or three separate heads that perform the mixing function independently before the flows are combined in a single discharge chamber. The ’128 Patent comes closest to the ’335 Patent in disclosing an L-shape device. The Patent’s principles are fairly simple, in that two straight-line mixing heads are joined at a 90° angle in order to better mix the polyurethane before it enters the mold. The key question is whether the location of the inlet nozzles and the short mixing chamber in the '335 Patent are true distinctions from the prior art. See Tr. 1778. Claim 1 of the ’335 Patent states the inlet nozzles “open into said mixing chamber close to said exit port means,” while Claim 6 states the inlet nozzles are “positioned in close proximity to said duct inlet so that mixing action initiated in said mixing chamber continues in said discharge duct,” (Emphasis added). By locating the inlet nozzles “close” to the discharge duct, thereby limiting the length of the mixing chamber, two important improvements over the prior mixing heads occur: first, the piston’s stroke and size are reduced, lessening the transient phenomenon. Second, the turbulent flow in the mixing chamber, caused in part by the small diameter of that chamber, continues as the polyurethane rebounds off the far wall of the discharge chamber because there is residual kinetic energy in the flow from the initial impingement. Additional mixing occurs at the top of the discharge chamber because of the L-shape design, while the chamber’s greater diameter permits a rapid dampening of the turbulent flow. PX 1, col. 4, lines 29-45. Powerful testimony at trial on the effect of the short mixing chamber and placement of the inlet nozzles a short distance from the discharge chamber came from Mr. Fritz Schneider, an employee of both KMC and KMAG. Mr. Schneider, who holds two patents for high-pressure mixing heads, has been intimately involved in polyurethane equipment manufacturing at least since 1971, in both design and marketing capacities. Tr. 1309-11. On cross-examination, the following exchange took place: Q. Okay. Isn’t it a fact that Afros ’335 patented mixing head has important advantages which are in no way obtainable with a Krauss-Maffei straight-line mixing head? A. In free pour applications, in most cases, yes. Q. And also in the advantage — in the ability to be used in a closed mold without an after mixer. Isn’t that a significant advantage as well? A. There are cases where it’s an advantage to have an L-shaped head, whether it’s Krauss-Maffei or Afros. Q. Even when you have got a closed mold? A. Yes. Tr. 1412. At the end of his testimony, the Court questioned Mr. Schneider further about the possible advantages of the ’335 Patent’s short mixing chamber: THE COURT: You said as I recall, I’m paraphrasing you, that the principle of the turbulence was the same in the middle figure as in the lower figure? [See infra, note 27, for figures.] THE WITNESS: Yes, the principle is the same. THE COURT: And that it [the location of the inlet nozzles] is commonly known? THE WITNESS: Yes. THE COURT: If it was so commonly known, do you know why people weren’t using the shorter mixing chamber and having it [inlet nozzles] close to the discharge cylinder? THE WITNESS: That is a very good question. It was at the time RIM machines were being developed. I would say hundreds of people were working on the same problems on the same principles. That is very often said af-terwards, you wonder why people don’t come up earlier on simple solutions. It is much more difficult to find a simple solution for a problem than a difficult one. You know, you mentioned, for example, the Baumgarten patent. You wonder, people were working on the same problem, very, very complicated. This is more discussion. I’m sorry. Maybe I shouldn’t do that. Tr. 1469-1469A (emphasis added). Mr. Schneider’s testimony is both highly revealing and exceptionally credible on the advantages of the ’335 Patent over previous mixing heads because he sought to overcome the same problems. The mixing chamber design, incorporating a short mixing chamber with the inlet nozzles located close to the discharge chamber, allows for use in open pour molds, which prior straightline mixing heads generally did not. This achievement was relatively simple, in hindsight, but as Mr. Schneider points out, simple solutions are often hard to discover. This evidence is probative of the differences between the prior art and the claimed invention. The expert testimony concerning the length of the mixing chamber in an L-shape head is in some agreement concerning the optimum length of a mixing chamber and location of the inlet nozzles. Plaintiff’s expert, Dr. Charles Garris, stated that the turbulent flow will diminish dramatically within two diameters of the point of impingement. Tr. 758. Dr. Macosko, defendant’s expert, stated a length-diameter ratio in the mixing chamber of less than one would prohibit the establishment of a laminar flow within the mixing chamber. Tr. 1591. He stated the turbulent flow will last between one and five diameters, depending on the velocity of the flow. Tr. 1592. Both experts agree that the movement around the 90° bend in the chambers will effect some additional mixing in the polyurethane flow, although the degree of mixing will vary depending on whether the flow is turbulent or laminar at that point. None of the prior art patents disclose a particular location for the inlet nozzles or the preferred length of the mixing chamber. Tr. 1403-1404. Reference to the patent drawings shows the length-diameter ratio for the ’128 Patent as 3.8:1, and for the Upjohn ’299 Patent as 7.3:1; the ’335 Patent drawings show a .5:1 ratio. Although patent figures are not necessarily drawn to scale and are for illustration purposes, they are helpful in this case in determining whether the prior art included the specific location of the inlet nozzles within its teaching. The Court concludes that the prior art does not relate the location of the inlet nozzles, as “close” to the discharge chamber, to the effect on mixing achieved in the discharge chamber. 3. Ordinary Skill in the Art In Environmental Designs, Ltd. v. Union Oil Co., 713 F.2d 693 (Fed.Cir.1983), cert. denied, 464 U.S. 1043, 104 S.Ct. 709, 79 L.Ed.2d 173 (1984), the Court of Appeals for the Federal Circuit listed the factors to be considered in ascertaining the hypothetical ordinary level of skill in the art: 1) the educational level of the inventor; 2) type of problems encountered in the art; 3) prior art solutions to those problems; 4) rapidity with which innovations are made; 5) sophistication of the technology; and 6) educational level of active workers in the field. Id. at 696. The Circuit Court noted that some factors may predominate in relation to a particular field. Id. at 696-97. The finding on ordinary skill in the art protects the integrity of the law of patents by focusing the Court’s attention away from a hindsight analysis and toward the state of the art prior to the issuance of the patent. “[A] person of ordinary skill in the art is also presumed to be one who thinks along the line of conventional wisdom in the art and is not one who undertakes to inno-vate_” Standard Oil Co. v. American Cyanamid Co., 774 F.2d 448, 455 (Fed.Cir.1985). The developmental period for high-pressure mixing heads, specifically from the straight-line to the L-shape design of the ’335 Patent, was roughly 1970 to 1977. The ordinary skill in the art did not change in any significant manner throughout this period. Tr. 566-67. Plaintiff’s expert, Dr. Jack Buist, stated the hypothetical person of ordinary skill would be a composite of a team. Tr. 567-68. The person would have a bachelor of science degree in mechanical engineering, some background in chemistry, and experience in the polyurethane manufacturing industry. Tr. 568-69. In addition, a member of the team that forms the composite would be a technician with expertise in engineering design. Tr. 568-69. Dr. Buist also testified the person would not have experience or training in fluid dynamics. Tr. 566. Dr. Macosko testified the person of ordinary skill would be aware of the '515 Patent’s teaching, have a bachelor of science in mechanical engineering with five years experience in mechanical design, and an understanding of fluid dynamics. Tr. 1701-1702. The problems facing the polyurethane industry in the relevant period involved questions of both mechanical design and fluid dynamics. The transient phenomenon presented difficulties in designing mixing heads, while the need to combine well-mixed polyurethane delivered in a splash-free pour presented fluid dynamics problems. The Court finds that, in addition to the consensus between the experts that a person skilled in the art would have a bachelor of science in mechanical engineering with experience in mechanical design, the level of ordinary skill included some level of knowledge and experience in fluid dynamics. The trend toward L-shape devices and the speed with which those developments occurred demonstrates the need for an ordinary person skilled in the art to possess some understanding of fluid dynamics if they were to participate in the field. The skill level was not great, but at minimum the person had a working knowledge of the effect of a particular design on the mixing and flow of polyurethane through the mixing head and into a mold. 4. Secondary Evidence The so-called “secondary considerations” identified in John Deere are of primary importance in determining obviousness, and the Court must always consider any external “real world” evidence that relates to the patent. Cable Electric Products, Inc. v. Genmark, Inc., 770 F.2d 1015, 1026 (Fed.Cir.1985); Vandenberg, 740 F.2d at 1567; see Stratoflex, Inc., 713 F.2d at 1538 (secondary considerations may be most probative and cogent evidence of nonobviousness). The question of what weight the evidence is entitled to depends “on its nature and its relationship to the merits of the invention.” W.L. Gore & Associates v. Garlock, Inc., 721 F.2d 1540, 1555 (Fed.Cir.1983), cert. denied, 469 U.S. 851, 105 S.Ct. 172, 83 L.Ed.2d 107 (1984). Plaintiff presented evidence of the commercial success of its FPL heads as an indicia of the nonobviousness of the ’335 Patent. “Commercial success is, of course, a strong factor favoring nonobviousness.” Akzo N.V. v. United States International Trade Commission, 808 F.2d at 1481. In considering whether a patent has achieved commercial success, the Court of Appeals for the Federal Circuit has instructed that “[a] nexus is required between the merits of the claimed invention and the evidence offered, if that evidence is to be given substantial weight en route to a conclusion on the obviousness issue.” Stratoflex, 713 F.2d at 1539. Defendant argues plaintiff’s raw sales figures do not establish the nexus between the patent and any inference of commercial success, and that plaintiff fails to place those figures in the context of overall market share from which meaningful inferences can be drawn. Plaintiff introduced figures for the years 1977 through 1986 for its worldwide unit sales of both straight-line and FPL mixing heads. Plaintiff also broke down the sales for the United States through Cannon, U.S.A., Inc. (“Cannon”), its exclusive American distributor. PX 31, 32. Annual worldwide FPL sales went from zero to 211 in the first three years, almost doubling each year. In the period 1980-84, total sales were well over 200 per year, and reached 356 and 362 in 1985-86. During the same period, worldwide sales of Afros straight-line mixing heads declined from over 200 in 1977 to 46 in 1985. PX 31. In the United States, where plaintiffs evidence is for the period 1978-86, unit sales show a similar but less dramatic growth, reaching 63 in 1986. PX 32. Afros’ sales, both worldwide and in the United States, have been further broken down into those sales only of the mixing head unit, and sales that include the Afros dispensing system. The latter system includes the mixing head, the chemical pumps and tanks, the frame, and other peripherals. Tr. 648-49. The system also includes in some sales the robotics for injecting the polyurethane in form molds automatically. Id. According to the testimony of Roy Brooks Hammer, president of Cannon, the overall increase in gross income from FPL sales is attributable in part to the sales of equipment accompanying the mixing head. Tr. 639-40. Defendant argues that the real basis for Afros’ success has been the application of the ’515 recirculation grooves in the FPL mixing head. In effect, KMC reiterates its nexus argument, asserting that the FPL head sales are not clearly linked to the invention in the ’335 Patent. However, aside from a reference at trial by counsel to that effect, Tr. 643, KMC presented no evidence beyond its unit sales to show the effect of the recirculation grooves on the commercial success issue. Defendant’s overall sales of the UL mixing heads since 1981 total 40, as compared to nearly 200 for Afros. Compare Pre-Trial Stipulations, Dkt. 127, H 54 with PX 32. Plaintiff’s sales data does not, standing alone, establish commercial success because it lacks the requisite nexus to the claimed invention. Cable Electric Products, Inc., 770 F.2d 1015 at 1027; Vandenberg, 740 F.2d at 1565. Two additional items do, however, supply the necessary link to permit the Court to consider the devices commercially successful because of the ’335 Patent’s invention. First, both Dr. Buist and Mr. Hammer testified the polyurethane molded product market has remained static almost throughout the period in which Afros marketed the FPL heads. Tr. 583, 640. The growth in FPL sales exceeds that of the market’s demand for polyurethane. Tr. 585. Second, the Afros head has replaced a number of competing heads, and importantly it has been substituted for straight-line heads. Tr. 592. The FPL head’s capacity to produce polyurethane molded items that straight-line heads could not, see Tr. 590, 636-37, is highly relevant evidence that the commercial success is directly attributable to the ’335 Patent, and not the recirculation grooves from the ’515 Patent found in most mixing heads. Defendant did not seek to impeach or rebut plaintiffs testimony on this question, and the Court credits the witnesses’ testimony that the FPL heads achieved strong commercial success attributable primarily to the ’335 Patent. Mr. Schneider’s testimony concerning the ’335 Patent’s “simple” solution to a persistent industry problem is also powerful objective evidence. See supra, text at 1416-1417. A patent responsive to an industry’s need, where a number of competitors have sought solutions to the problem without success, supports a finding of non-obviousness. See Akzo N.V. v. United States International Trade Commission, 808 F.2d at 1481 (solution to longfelt need immediately recognized by practitioners as “a remarkable advancement” evidence of nonobviousness). Mr. Schneider’s testimony also supports Dr. Fiorentini’s statement that his design achieved an unexpected result by both limiting the transient phenomenon and achieving better mixing while a laminar flow was created at the point of injection into the mold. Tr. 188-90. Mr. Schneider, who is clearly well versed in the art, could not explain why the solution to the problem was missed. Moreover, the prior art does not show a short mixing chamber in combination with the inlet nozzles close to the discharge chamber as a viable answer. By locating the inlet nozzles within a relatively short distance from the discharge chamber, the length of the mixing chamber can be reduced because the area to be cleaned after a shot and the stroke of the mixing piston are reduced. The evidence of an unexpected advantage achieved by the ’335 Patent is “strong support for a conclusion of nonobviousness.” Lindemann Maschinenfabrik, 730 F.2d at 1461. 5. Conclusion In its opinion on remand in Panduit Corp., the Court of Appeals for the Federal Circuit made abundantly clear the legal standard for analyzing the issue of obviousness: To reach a proper conclusion under § 103, the decisionmaker must step backward in time into the shoes worn by the “person [having ordinary skill in the art]” when the invention was unknown and just before it was made. In light of all the evidence, the decisionmaker must then determine whether the patent challenger has convincingly established, 35 U.S.C. § 282, that the claimed invention as a whole would have been obvious at that time to that person. 35 U.S.C. § 103. The answer to that question partakes more of the nature of law than of fact.... 810 F.2d at 1566 (emphasis in original). Both the claimed invention and the prior art must be considered as a whole. Lindemann Maschinenfabrik, 730 F.2d at 1462; Hybritech Inc., 802 F.2d at 1383 (Fed.Cir.1986). As the Circuit Court has noted, “Obviousness cannot be established by combining the teachings of the prior art to produce the claimed invention, absent some teaching, suggestion or incentive supporting the combination.” In re Geiger, 815 F.2d 686, 688 (Fed.Cir.1987) (citing ACS Hospital Systems, Inc. v. Montefiore Hospital, 732 F.2d 1572, 1577 (Fed.Cir.1984)). In reviewing the ’335 Patent in light of the prior art, the fundamental distinction is the location of the inlet nozzles “close to” and “in close proximity to” the discharge chamber, permitting a design that uses a short mixing chamber. The prior art does not specify the inlet nozzle location, but it does disclose the other major facets of the ’335 Patent: recirculation grooves, fluid communication between the mixing and discharge chambers, a discharge chamber with a greater cross-sectional area than the mixing chamber, and the L-shape configuration. None of the prior art, including the ’128 Patent, teaches the placement of the inlet nozzles at any particular distance from the discharge chamber, permitting impingement in a short mixing chamber as affecting the component mixing or transition to a laminar flow. On this point, the prior art discloses at most a mixing chamber as narrow as possible, considering the volume necessary for production, in order to effectuate component impingement. The testimony conflicts as to whether a person of ordinary skill in the art would understand whether the L-shape and its relation to the location of the inlet nozzles in permitting use of a short mixing chamber would have an effect on mixing. Mr. Schneider, who holds patents on high-pressure mixing heads and has long experience in the industry, testified the ’128 and ’335 Patents, because of their L-shape, are based on the same principle and have similar teachings on impingement mixing. Tr. 1449. These statements conflict with his later response that the ’335 Patent’s solution eluded the industry because of its very simplicity. Tr. 1469. It is incongruous to assert the prior art made obvious an invention while also stating the ’335 Patent effected an advance in polyurethane mixing head technology. The Court has previously credited Mr. Schneider’s testimony concerning the advantage of the ’335 Patent design, and therefore must reject his conflicting testimony on the disclosures of the prior art. The opinions of the expert witnesses concerning the fluid dynamics of the flow in the mixing heads conflicts as to whether the prior art reveals the claims of the ’335 Patent. Dr. Macosko stated the need to create “back pressure” in the mixing chamber can be effected through an L-shape, Tr. 1364, but he did not see any further beneficial effect from the short mixing chamber design and location of the inlet nozzles on mixing in the discharge chamber. Tr. 1585-86. His testimony was based on knowledge of general principles of fluid dynamics and not from experience working with L-shape heads. Plaintiff’s expert, Dr. Garris, stated the L-shape configuration, when combined with a short mixing chamber, “effectively utilized” the energy created through placement of the inlet nozzles close to the discharge chamber in the area of impingement to promote better mixing. Tr. 758-59. He noted that when first confronted with the ’335 design, he did not know whether the short chamber would aid the component mixing. After performing a visualization test, Dr. Garris saw the effect of the FPL head’s design on mixing. Tr. 757-58. Dr. Garris’ field of expertise is in fluid dynamics and reading blueprints, not as an expert in polyurethane mixing heads. Tr. 696-97. Defendant’s expert testified that the value of an L-shape design is clearly known, and the length of the mixing chamber is dependent on the volume of the shot, but the configuration does not otherwise affect mixing. Plaintiff’s expert states the ’335 Patent creates a better mix, and the value of a short mixing chamber through the location of the inlet nozzles had not been revealed prior to the ’335 Patent. Both witnesses lack expertise in the polyurethane mixing head industry, which limits the weight of their testimony. They are qualified as experts in fluid dynamics, and the Court has considered their testimony in light of that limitation. Professor Macos-ko’s testimony conflicts with the statements of Mr. Schneider concerning the effect of the location of the inlet nozzles because defendant’s expert does not account for the additional mixing caused by the short length of the mixing chamber and placement of the inlet nozzles combined with the L-shape design. In weighing the experts’ testimony on whether the mixing chamber length affects the component mixing, objective