Nicofibers, Inc. v. Reichhold Chemicals, Inc.

OPINION AND ORDER DUNCAN, District Judge. Introduction This is an action for patent infringement. The subject matter of the action is a patent (“the Miller patent”) which has been assigned to plaintiff Nicofibers, Inc. The Miller patent describes a process for making multi-layer glass fiber mats. Plaintiff Nicofibers seeks a judgment that the process used by defendant Reich-hold Chemicals, Inc., for producing multilayer glass fiber mats infringes the Miller patent. Plaintiff seeks an injunction against further infringement, damages for past infringement, treble damages for defendant’s alleged “wilful and deliberate” infringement, and attorneys’ fees. Defendant Reichhold counterclaims, seeking a declaratory judgment that the Miller patent is invalid and not infringed. This Court has personal jurisdiction over the parties. Subject matter jurisdiction is based on 28 U.S.C. §§ 1338(a), 2201 and *498 2202. Venue is proper under 28 U.S.C. § 1400(b). This matter was tried to the Court. The Court has before it the evidence adduced at trial and the briefs and argument of the parties, and it has benefited by personal inspection of the processes in dispute at the plant sites of both the plaintiff and defendant. What follow are the Court’s findings of fact and conclusions of law pursuant to Rule 52(a), Fed.R.Civ.P. The Parties The plaintiff, Nicofibers, Inc., is a Delaware corporation with its principal place of business in Shawnee, Ohio. Defendant Reichhold Chemicals, Inc. is a Delaware corporation with its principal place of business in White Plains, New York. Reichhold’s Glass Fiber Division, with which this suit is principally concerned, operates a manufacturing plant in Bremen, Ohio. Nicofibers, Inc. was founded in 1972 by Nicolet Industries. Several employees of Reichhold, including the patentee William Miller, left Reichhold’s Glass Fiber Division to join Nicolet and were instrumental in initiating the Nicofibers operation at Shawnee, which now competes with Reichhold in the production of glass fiber mats. Wellspring of the Miller Patent: The Modigliani Process Both Nicofibers and Reichhold manufacture glass fiber mats by a method known as the “Modigliani process.” Named after its inventor, Piero Modigliani, this process has been used to produce glass mats for more than 40 years. Essentially the process works as follows: A furnace containing a supply of broken glass traverses back and forth over a large rotating drum, parallel to the drum’s axis of rotation. As the glass melts in the furnace it is drawn through small orifices in the bottom of the furnace, emerging as fine fibers or filaments of molten glass. These fibers are attached to the drum, and as it rotates, the fibers are wound into layers upon the drum, much as fishing line is wound upon a reel. The traversing speed of the furnace and the rotating speed of the drum are adjusted so that the glass fibers in each layer formed on the drum cross at acute angles with respect to fibers of adjacent layers. The layers are allowed to accumulate as the filaments are wound on the drum to form a condensed glass fiber mat, a step which takes from about 90 minutes to several hours to complete, depending upon many factors including the speed and size of the drum and the thickness of the mat being spun. During the winding of the fibers on the drum, the glass fibers are loosely bound and relatively weak. In order to bind them together into a cohesive mat a polyester resin, or “binder solution,” is uniformly applied to the layers periodically by means of an automatic spray gun. When the mat has been built up to the desired thickness, the drum is stopped and the condensed mat is cut off the drum parallel to its axis of rotation. The mat is then placed on a table and stretched (“expanded”) to increase the porosity and reduce the density of the product. The expanded mat is then conveyed to a curing oven in which hot air is blown through the mat to cure the binder solution, and finally the expanded, cured mat is rolled up and packaged for shipment to the customer. The Modigliani process is used by both Nicofibers and Reichhold to produce single layer “surfacing mat” and “reinforcing mat.” These are used as components in the production of a wide variety of reinforced plastic items such as cafeteria and TV trays, electrical panel board, welding helmets, carrying cases for office equipment, and motorcycle and automobile body parts. Reichhold also produces glass fiber “media” for air filtration products and a glass fiber mat product known as “veil” which performs essentially the same function as surfacing mat except that it is made to conform more easily to complex shapes and curves. Surfacing mat is manufactured in varying degrees of thickness from 10 “mils” (one mil is .001 inch) to approximately 30 mils. Reinforcing mat, which consists of thin sur *499 facing mat layers sandwiching a swirled reinforced mat, is often made thicker. A ten mil mat is as thin a mat as can be made successfully on a commercial basis, according to Nicofibers’ president, John K. Whitacre. A surfacing mat, because of its inherent “layered” nature, can be split manually to produce mats thinner than 10 mils; however it is a “tedious” and unwieldy process. Bane of the Business: Delamination As the Court has already observed, a critical step in producing a cohesive glass fiber mat is the application of the resin solution which serves to bind the fibers together. Because the solution “migrates” or soaks into the spinning layers it is sprayed intermittently, in fixed “on” and “off” periods, rather than continuously, onto the drum. For example, in the Modigliani process used by Nicofibers, the binder spray is turned “on” for a period of one minute, then “off” for a period of 1.1 minutes, and this cycle is continuously repeated. Reichhold uses an “on” cycle of 47 seconds every 2.5 minutes. By contrast with these planned and preprogrammed on-off cycles, it not infrequently occurs in the Modigliani process that for various reasons the application of the binder spray is accidentally interrupted and remains “off.” In this situation the glass fibers continue to be wound into layers on the spinning drum, but without the resin spray there is nothing to bind them together. The result, if the interruption of the spray continues for any substantial period of time, is a layer of dry fibers which, without cohesion, falls off of the drum or falls away from the rest of the mat, generally in a tangled mess, during the expansion and curing process. This phenomenon is known as “delamination.” Commencing with the introduction of automatic spray equipment in the early 1960’s delamination has become a substantial and well-known problem throughout the industry, resulting most frequently from mechanical failures in, or clogging of, the spray equipment. Every witness called by the parties to testify in the case testified that he was aware of the phenomenon of delamination prior to 1973, and knew that it was caused by the inadvertent spinning of a dry fiber layer into the mat. Delamination is generally considered a defect and is a major source of rejected material in the industry. The Miller Patent On March 25, 1975, United States Patent No. 3,873,291 was issued to William R. Miller. 1 Essentially, the Miller patent de *500 scribes a process 2 by which several glass fiber mats can be produced on the spinning drum, one on top of the other, without stopping the drum to cut off the individual mats. The Miller process permits these mats to be spun sequentially and in such a fashion that they can be separated after they are removed from the drum, stretched and cured. The Miller process works as follows: Initially a mat is spun on the drum using the Modigliani process. When this mat reaches its desired weight and thickness, rather than stopping the furnace and drum to cut this single mat from the drum, the Miller process allows glass fibers to continue to be wound on top of the first mat. For a predetermined and controlled period of time, no resin binder solution is sprayed on these fibers. Without the binder solution these fibers accumulate as a dry layer over the first mat. At the end of the “dry” period the spray is re-started and a second mat is built up using the normal Modigliani process. After this mat is completed the spray gun is again shut off for the predetermined period and another dry layer is spun onto the drum, followed by another “Modigliani” mat. This alternating process continues until the desired number of mats-separated by dry layers-is formed. The product of the Miller process-trademarked “Layer-Pak” by Nicofibers-is then cut in its condensed form from the drum, stretched, and processed through the curing oven. The end result is a product which looks like a thick single mat, but which in fact consists of a number of uniform and separable individual mats. This multi-layered product is rolled and shipped to customers who then separate it according to their production needs. Scrutiny of the Miller process discloses that it in effect harnesses the well-known phenomenon of delamination and makes it a useful tool for producing multi-layer mats. Indeed, Miller characterized his first attempts to produce the multi-layer product as “programmed delamination experiments.” Miller testified: Q. Why did you use the term delamination for this experiment? A. At that point in time, that is essentially what we were trying to do. We were trying to separate the mat. Q. Is delaminated mat the same thing as what you were trying to do? [Objection overruled.] A. Delaminated mat, like I said, is a reject, unless it is done in a controlled manner. Q. What further experiments did you run? A. We ran another experiment to determine-to refine it even further, the delamination. T-I-81-82, 83. From the manufacturing standpoint, assuming customers who are willing or able to *501 accept glass fiber mats in the multi-layer format, the Miller process does provide a number of advantages over the previous method of production. 3 For example, the cost of producing five glass fiber mats using the Miller process is less than the cost of producing the same five mats individually with the straight Modigliani process. With the Miller process the spinning drum and furnace can be kept in operation continuously while the five mats are spun, rather than being stopped and then restarted after each mat is made and cut off. Likewise, when a multi-layer mat is stretched and put through the curing oven, labor and manufacturing resources are utilized more efficiently than if the mats were processed individually. Contamination problems are also reduced since the inner mats are protected from outside contact until they are separated for actual use. Finally, transportation and storage costs are conserved since the rolls of multi-layer material are denser and allow a more efficient use of shipping and inventory space. Development of the Patent The key element of the Miller process is the incorporation of the controlled binder spray shut-off, and the resulting dry fiber layer, into the Modigliani process. It is this procedure which permits the layered mats to be produced sequentially, without interruption, and then later to be separated into individual mats. Additionally, Miller determined the timing of the binder spray shutoff required to produce the optimum product. Miller testified that he conceived the idea of making a multi-layer mat in early March 1973 as a product with which to generate possible new business or “crack the market” for the fledgling Nicofibers operation. In order to determine whether, and how, a multi-layer mat could be produced, Miller set up his “programmed delamination experiment.” Essentially this involved spinning a series of mats and shutting off the binder spray between each mat for a progressively longer period of time. Thus, in the first experiment, Miller ordered four mats to be spun with successive shut-off periods of 4 minutes, 20 seconds; 7 minutes; and 9 minutes, 40 seconds. According to Miller the 4*/3 minute period was too short to produce separation of the mats, while the 9% minute period produced “excessive separation.” The 7 minute period was “pretty good.” Further experiments over the succeeding few days refined the time period until Miller determined that a period of approximately 8% minutes produced the optimum separation effect. Nicofibers began offering the multi-layer mat to customers as a viable commercial product at least by late April 1973. “Layer-Pak” was announced to the industry generally in a news release dated May 18, 1973. Reichhold’s Multi-Layer Product Development As stated, Miller initially developed his multi-layer product in March 1973, and Nicofibers was selling it commercially by May 1973. At some time during this period, “early in 1973,” according to defendant’s sales manager, Eugene Carney, Reichhold learned of the Nicofibers product through several of its customers including one in Ashtabula, Ohio, the Molded Fiberglas Company (“Molded”). Molded showed Carney a sample of the Nicofibers multi-layer product and asked if Reichhold could duplicate it. Carney said that they could, although it appears from his testimony that he has not sure how it would be done. The evidence is clear that Reichhold had no interest in manufacturing a multi-layer product until they learned of the Nicofibers product. Carney returned to Reichhold and presented the problem to them. He testified as follows: *502 Q. Now, after your visit to Molded Fiberglas did you do anything in respect to their inquiry about whether [Reichhold] could make multilayer mat? A. I returned to my plant and told my manufacturing people that I would like to have a product like that. Q. Were they able to supply you with the product? A. Yes. The evidence shows that Reichhold obtained samples of the Nicofibers product, as early as April 1973, through several of Reichhold’s customers, apparently for analysis purposes. Although analysis was performed by Reichhold, Robert Lovell, Reichhold’s operations manager at the Bremen plant, testified that when they first learned of the Nicofibers product Reichhold did not refer the problem to their research and development department, bypassing it instead to go directly to the manufacturing department and spinning manager, William Rodgers. Rodgers testified that when the problem of making a multi-layer product was first presented to him he was only given a brief description of the Nicofibers product, and did not actually see a sample. Rodgers testified as follows: Q. Had someone prior to 1973 come to you and said, “Bill, would you spin me a multiple, one multiple-layer mat,” would you have been able? A. They had not asked me. Q. If someone had come and asked you that, would you have known how to do it? A. Yes. Q. How would you have done it? A. By using the spray system, turning the spray off a period of time, and making the mat delaminate is actually all it is, controlled delamination. Q. Mr. Rodgers, in 1973 did someone ask you whether you could make a multiple-layer mat on your spinning machine? A. Yes. Q. What was your response to that? A. I answered this would be no problem, that I might have to make a step or two to get just as what we-what I wanted. I knew I could do it. Q. Do you recall the circumstances, where and when you were asked that? A. I cannot remember if it was a special meeting called, or if it was a regular morning meeting we have every morning at ten. I cannot recall this. I know it was at a meeting I was asked this question. Q. Do you recall who asked you the question? A. No. Q. Did you take any action in respect to this request to manufacture a multimat? A. Yes, I started. I only made two-layer mat to begin with. I turned the spray off in the two layers only two minutes at a time. After expansion we found this was not a successful thing. Then we increased to four minutes off during the middle. This was better, yet not what we were trying to reach. Q. What did you next do? A. We went to six minutes. The same thing, it was not successful, and we felt it was not what it should be because with facilities of spinner, what have you-we increased it to eight. By continuing working with eight minutes off it seemed to work, both ends of it plus off spraying making successful layered mat. Q. From that time did you continue to use eight minutes as off time? A. Yes, if we are required to make it now we use eight minutes. T-II-98-100. The Court credits this testimony of Mr. Rodgers. Although he could not supply the exact date when this development occurred, *503 evidence at trial demonstrated that Reich-hold manufactured multi-layer mat under the “Rodgers process” at least as early as July 1973, and that Reichhold was shipping its new “Stratamat” to customers in August and September 1973, 4 more than six months prior to Miller’s patent application, and a year and a half prior to the issuance of the Miller patent. After its initial development of a multilayer product, Reichhold continued to analyze Nicofibers’ Layer-Pak. In November of 1974 an analyst reported that Nicofibers was using a “powdery release agent that looks very much like talc” between the layers. Prompted by this report, Reichhold, in March of 1975, undertook experiments involving the use of talc as a release agent in its multi-layer process. Nicofibers did in fact attempt to use a talc release agent, but not until three or four years after initial multi-layer development in 1973. Reichhold continued to manufacture Stratamat in 1973 and 1974; however its sales have been relatively low since then, totalling only about $60,000 through the filing date of this lawsuit. Sales manager Carney testified that he “[does] not actively pursue multilayer mat sales.” Developments at United States Gypsum Co. The Court heard from witness Thomas Kessinger, a former mechanical engineer for the United States Gypsum Co. (USG) in Wabash, Indiana. 5 Kessinger held various positions with USG from 1964 to 1973 including project engineer, quality superintendent and production superintendent. In 1973 he left USG to become plant superintendent at Reichhold’s Bremen facility. USG produces glass fiber filtration mats, or “filter media” using the Modigliani process. The production process at USG is virtually identical to that of Reichhold and Nicofibers except that USG uses a different binder resin and sprays it continuously on the mat rather than in cycles. Delamination was a problem to USG but not a “major” one, according to Kessinger. He testified that USG only experienced delamination problems during a power failure. During such an occurrence the spray gun and drive motors would stop operating, but the large drum would continue rotating for several minutes due to its own inertia. Because the motor drive shaft of the drum also drove the traversing furnace, fibers would continue to be wound across the drum from the furnace so long as the drum kept spinning. Because the spray gun was cut off, however, these fibers would go on dry, resulting in unwanted separation or delamination. In approximately 1968 a USG foreman working for Kessinger conceived the idea of spinning a two-layer mat on the drum. Kessinger testified: A. He [the foreman] took the media and broke it into two pieces. He said, “All we have to do is spin the mat like this, and we can run two mats in the oven at the same time and double our oven capacity.” Q. Did you take any action then in respect to this suggestion he had made? A. Yes, we began to try to run those mats through our oven. Q. In terms of spinning a mat, did you spin a mat a different way then in order to accomplish the ability to run the mat-two mats through the oven at one time? A. AH they did is go back to the spinning department. Our entire spinning-let’s say the drum and the resin and the furnace cycling was all on one circuit, so when our man hit a kill button, he killed everything, so all he did when we were between mats, he went back and hit the button, let it just stop in the cycle for *504 several minutes, hit the start button, and would go right back through the process. Q. As a result of that, you then would spin two mats on the same drum? A. Yes. Q. Mr. Kessinger, did you go through an R and D program or any research to develop this process to make two mats on a single drum? A. No, the foreman who walked into my office, he had seen it many, many times. He knew how to do it. T — III—11, 12 (emphasis added). USG discontinued the multi-layer process after a week or two because it did not produce mats of uniform “fluffiness,” a characteristic which is, by contrast to surfacing and reinforcing mat, requiring for filtration mats. The Court credits the testimony of Mr. Kessinger and finds it persuasive that the theory which William Miller used to produce his plurality of separable concentric glass fiber mats beginning in 1973 was essentially the same theory used at USG several years earlier, despite the mechanical differences in its application. Discussion of Law There are three essential elements of patent validity: novelty, utility and nonobviousness, codified in 35 U.S.C. §§ 101-103. United States v. Adams, 383 U.S. 39, 48, 86 5. Ct. 708, 712-713, 15 L.Ed.2d 572 (1966). Hanson v. Alpine Valley Ski Area, Inc., 611 F.2d 156, 159 (6th Cir. 1979); Reynolds Metals Co. v. Acorn Bldg. Components, Inc., 548 F.2d 155, 159 (6th Cir. 1977). Defendant Reichhold’s principal contention is that plaintiff’s patent is invalid because it is “obvious” within the meaning of 35 U.S.C. § 103. In analyzing the validity of any patent the Court must start with the proposition that every patent issued by the Patent Office “carries with it a statutory presumption of validity under 35 U.S.C. § 282.” Hanson, supra, 611 F.2d at 159; Reynolds Metals Co., supra, 548 F.2d at 160. At least this presumption applies where the pertinent prior art is considered by the Patent Office. Where the prior art is not so considered the presumption is weakened. Smith v. ACME General Corp., 614 F.2d 1086 at 1090-1091 (6th Cir. 1980); Hanson, supra, 611 F.2d at 159. In the present case the prior art has been considered by the Patent Office, 6 and the Court therefore accords the patent the statutory presumption. The § 282 presumption has no independent evidentiary value, however; its purpose is merely to allocate the burden of proof of invalidity to the party asserting it. Eltra Corp. v. Basic, Inc., 599 F.2d 745, 750 (6th Cir. 1979), cert, denied, 444 U.S. 942, 100 S.Ct. 297, 62 L.Ed.2d 308. Reynolds Metals Co., supra, 548 F.2d at 160. Further, the Court finds in this case no “unusual factual circumstances,” Dickstein v. Seventy Corp., 522 F.2d 1294, 1296 (6th Cir. 1975), cert, denied, 423 U.S. 1055, 96 S.Ct. 787, 46 L.Ed. 644 (1976), which would require the application of a standard of proof stricter than a preponderance of the evidence test. See, Eltra Corp., supra, 599 F.2d at 750-51; Campbell v. Spectrum Automation Co., 513 F.2d 932, 935-37 (6th Cir. 1975). *505 As stated, Reichhold contends that the Miller patent is invalid because it is obvious. 35 U.S.C. § 103 states the nonobviousness requirement as follows: A patent may not be obtained though the invention is not identically disclosed or described as set forth in section 102 of this title, if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains. Patentability shall not be negatived by the manner in which the invention was made. The landmark guiding judicial inquiry into obviousness under § 103 is Graham v. John Deere Co., 383 U.S. 1, 86 S.Ct. 684, 15 L.Ed.2d 545 (1966). There the Supreme Court outlined the inquiry as follows: (1) [T]he scope and content of the prior art are to be determined; (2) Differences between the prior art and the claims at issue are to be ascertained; (3) And the level of ordinary skill in the pertinent art resolved. (4) Against this background, the obviousness or nonobviousness of the subject matter is determined. (5) Such secondary considerations as commercial success, long felt but unsolved needs, failure of others, etc., might be utilized to give light to the circumstances surrounding the origin of the subject matter sought to be patented. As indicia of obviousness or nonobviousness, these inquiries may have relevancy. 383 U.S. at 17, 86 S.Ct. at 693-694 (clauses separated and numbered for convenience). The issues contained in paragraphs (l)-(3) above, as well as the “secondary considerations” in paragraph (5), present questions of fact for the Court. The ultimate issue of obviousness, paragraph (4), is a question of law. Sakraida v. Ag. Pro., Inc., 425 U.S. 273, 280, 96 S.Ct. 1532, 1536, 47 L.Ed.2d 784 (1976); Eltra, supra, 599 F.2d at 751; Nickola v. Peterson, 580 F.2d 898, 911 (6th Cir. 1978); cert, denied, 440 U.S. 961, 99 S.Ct. 1504, 59 L.Ed.2d 774 (1979). The Court will now turn to these issues. Scope and Content of the Prior Art The parties do not dispute that the prior art relevant to the Miller patent is the Modigliani process, and the various patents covering it, for producing glass fiber mats. 7 That the Modigliani process is not itself a process for producing multiple-layer mats is of course not disputed. What the parties do dispute is the relevance of the phenomenon of delamination. The defendant claims that delamination is “prior art” which should be considered in determining obviousness. Plaintiff insists, on the other hand, that delamination is merely an uncontrollable defect in the Modigliani production process and should not be considered as prior art. The Court acknowledges, and finds, that delamination has been the bete-noir of the glass fiber mat industry, generally producing a product which must be rejected as unuseable. Thus, Reichhold’s argument that it had “inadvertently produced hundreds of separable mats prior to 1973 due to interruption in the binder spray caused by malfunctioning spray guns,” defendant’s Brief at 12, is somewhat misleading. Nevertheless, whether delamination can be considered prior “art” is not the point here. Delamination, art or not, is a phenomenon intimately connected with the Modigliani process. Every witness who appeared at trial testified about the delamination problem. It is abundantly clear that persons familiar with the prior art Modi *506 gliani process are also familiar with the causes and effects of delamination. Recently the United States Court of Appeals for the Sixth Circuit defined the boundaries of “prior art” as follows: Prior art need not be exactly the same in order to be applicable; differences between the prior art and the invention are permitted. That is to say that the mere existence of differences between the pri- or art and an invention does not establish the invention’s nonobviousness particularly where, as here, the prior art and the invention are substantially similar .... [W]e point out that the object and the accompanying claims as well as the inferences one skilled in the art would draw from such specific disclosures, properly constitute prior art. In fact, ‘‘the prior art is all the knowledge that would have been available to any person having ordinary skill in the art.” (Citing Steelcase, Inc. v. Delwood Furniture Co., 578 F.2d 74 (5th Cir. 1978), cert, denied, 440 U.S. 960, 99 S.Ct. 1503, 59 L.Ed.2d 774 (1979) (emphasis added). Smith v. ACME General Corp., supra, at 1093. Thus, when Congress included the requirement of nonobviousness in § 103 the would-be inventor became “chargeable with what he should have known, under the prior art, rather than merely with what was explicitly disclosed by the inventions of his predecessors in the art.” Allegheny Drop Forge Co. v. Portee, Inc., 370 F.Supp. 673, 676 (W.D.Pa.1974), aff’d., 541 F.2d 383 (3d Cir. 1976). Further, it has been stated: The “scope and content” of the prior art consists of prior patents, publications, products, and methods in use at the time the invention was made .... The pertinent prior art is that to which one can reasonably be expected to look for solutions to the problems which the patented device attempts to solve. (Emphasis added.) Fischer & Porter Co. v. Haskett, 354 F.Supp. 464, 477 (E.D.Pa.1973). In view of the above, the Court finds that the scope and content of the prior art consists of the Modigliani process and the patents covering it, above cited. Included as part of this prior art must be the inferences and knowledge which would be drawn from the Modigliani process by a person having ordinary skill in the art. Smith v. ACME General Corp., supra. Such inferences and knowledge would certainly encompass one of the most pervasive problems involved in successfully completing the process-delamination. Differences Between the Prior Art and the Claims at Issue The Miller patent contains nine claims (see footnote 1, supra), of which the principal claims are numbers 1 and 5. Claim 1, which is representative of the other claims, reads as follows: 1. Method of sequentially producing a plurality of glass fiber mats, comprising: a. continuously drawing glass filaments from a molten glass supply through a plurality of orifices; b. continuously winding the filaments about the periphery of a drum rotating about its axis; and c. intermittently applying a binder solution to the filaments being wound continuously on the drum, the binder solution being applied cyclically to the filaments at predetermined intervals during a first period of time sufficient in duration to form a first integral mat, the binder solution not being applied during a second period of time of predetermined duration such that an integral mat is not formed at a location, and the binder solution being applied cyclically to the filaments at predetermined intervals during a third period of time sufficient in duration to form an overlying second integral mat concentric with and separable from the first mat at said location where no binder solution is applied. (Emphasis added.) The Court finds that the quoted portion of this claim, down to that which is emphasized, describes the Modigliani process. As *507 has previously been discussed, the change worked by the Miller process-shown in the emphasized portion-is that of stopping the binder spray and allowing a layer of dry fiber to accumulate, facilitating separation of the adjacent mats. Thus, the difference between the claims contained in the Miller patent and the prior art is this additional step of interrupting the binder spray, within specific time parameters, to permit the production and separation of the multiple concentric mats. With regard to delamination, the difference between the Miller process and the “prior art” is most aptly captured in Mr. Miller’s own initial characterization of his process as “programmed delamination.” The difference lies in the word “programmed.” Miller simply took the known cause of delamination, isolated it, and programmed it to produce a viable commodity, multi-layer mat. Level of Ordinary Skill in the Art In making the determination of obviousness, the standard of judgment is not what would be obvious to a layman, “but rather what would be obvious to one ‘reasonably skilled in the applicable art.’ ” Dann v. Johnston, 425 U.S. 219, 229, 96 S.Ct. 1393, 47 L.Ed.2d 692 (1976). This has been called “a specialized reasonable man test for obviousness,” Burndy Gorp. v. Kearney-National, Inc., 466 F.Supp. 80, 88 (S.D.N.Y.1979). Determining the level of ordinary skill in the art is a two-step process. General Motors, Inc. v. Toyota Motor Co., Ltd., 467 F.Supp. 1142,1170-1171 (S.D.Ohio 1979) (“Toyota”): First, the body of people who devote regular and substantial effort to solving the type of problem that the alleged invention solves must be identified. Second, the level of skill of an ordinary member of that group must be determined. Conducting the first inquiry in light of the evidence produced at trial, the Court finds that the hypothetical person is a member of the body of people who have at least several years of work experience since the years 1960-61 in the glass fiber mat industry using the Modigliani process. Such persons would be familiar both with the mechanics and theory of the spinning process and with quality control in the production of glass fiber products made by the Modigliani process. Under the second inquiry the Court looks to a number of indicia. See Toyota, supra, 467 F.Supp. at 1171-1172. In particular it looks to evidence of general information and knowledge which would have been possessed by an ordinary practitioner in the group defined in step one, i. e., those devoting “regular and substantial effort to solving the type of problem that the alleged invention solves.” Here, educational background and depth and breadth of experience in the art are relevant. Further, evidence of work habits and procedures performed in the art is appropriately considered. Id. These inquiries permit the Court, [b]y producing evidence of the predispositions, motivations, methodologies and work routine of practitioners in the pertinent art ... to construct a hypothetical paradigm which can be used to identify obvious outcomes to problems. Toyota, supra, at 1172. The evidence introduced at trial suggests that the hypothetical person sought here is one intimately familiar with the Modigliani process, not necessarily through formal engineering training, but through experience working at and supervising the manufacture of glass fiber mat products by that process. The Court finds that such a person would be in either the manufacturing operations or quality control departments and would hold the position of foreman or supervisor, responsible for ensuring that production proceeded smoothly in the plant. This person would of course be familiar with the phenomenon of delamination. Obviousness Having made the preliminary factual determinations required by Graham v. John Deere Co., supra, the Court now turns to *508 the “ultimate question of obviousness.” Smith v. ACME General Corp., supra. The crux of this question has been stated by the Supreme Court as follows: [T]he “obviousness” test of § 103 is not one which turns on whether an invention is equivalent to some element in the prior art but rather whether the difference between the prior art and the subject matter in question “is a difference sufficient to render the claimed subject matter unobvious to one skilled in the applicable art Dann v. Johnston, 425 U.S. at 228, 96 S.Ct. at 1398 (citation omitted). In reaching its conclusion the Court notes particularly that “obviousness is measured not by considering what was obvious to actual artisans, but by considering whether a hypothetical person, having all of the prior art at hand, would have found the same solution when addressing himself to the same problem.” Republic Industries, Inc. v. Schlage Lock Co., 592 F.2d 963, 975 (7th Cir. 1979) (emphasis added). In the present case the Court concludes upon all the evidence that the solution to the problem addressed by William Miller in March of 1973 would have been, and in fact was, as obvious to others reasonably skilled in the art as it was to the patentee himself. For that reason the patent must fail. In the Court’s view, “once the problem was recognized, the place to find the solution became apparent.” Id. In particular, within a few months of the time that Miller’s multi-layer mat became available commercially, and more than a year prior to the issuance of the patent in dispute, defendant Reichhold was able to produce its own commercially viable multilayer mat using the process claimed by the patent. The Court is persuaded by the credible testimony of William Rodgers that once a multi-layer product was requested he simply went to his spinning department at Reichhold and produced it, virtually duplicating the process of Mr. Miller at Nicofibers. This is corroborated by the testimony of Thomas Kessinger, who was then Rodgers’ superior at Reichhold: Q. Did you relay such instructions [to produce multilayer mat] to Bill Rodgers? A. Yes. Q. Did they go out and manufacture multilayered mat? A. Yes. T — III—151 Likewise, Eugene Strohl testified with regard to a memorandum (defendant’s Exhibit “R”) he sent to Rodgers which stated as follows: 9/25/73 W. Rodgers Run two stratamats (5 layers) at nominal of 6% binder Nominal flow 26.4 grams Range 23.0 to 30.0 Strohl testified: Q. Will you briefly tell us what that document is? A. It is a note to Bill Rodgers to run two stratamats at six percent binder and I give him the nominal flow and range for flow. Q. Did you tell him how to manufacture the stratamats? A. No. Q. Why did you not tell him how to manufacture the stratamats? A. Because we already knew how at this time. Q. What was the date of that note? A. September 25, 1973. Kessinger did testify on cross-examination that Reichhold manufactured a number of flawed multi-lawyer mats early in its production, and that he thought the process was a “pain in the neck from a production standpoint.” Nevertheless, this cannot undercut the fact that the core of the Miller process-stopping the binder spray in order to spin a dry fiber layer to cause and facilitate separation of mats-was obvious to Reichhold’s production people as soon as the problem was presented to them. Thus, this *509 is not a case in which “experts, who devoted years of effort, failed to achieve an equivalent result” e. g., Bolkcom v Carborundum Co., 523 F.2d 492, 500-501 (6th Cir. 1975), cert, denied, 425 U.S. 951, 96 S.Ct. 1725, 48 L.Ed.2d 194 (1976), or in which experts, when told of the invention, did not believe it would work, e. g., United States v. Adams, supra, 383 U.S. at 43-44, 86 S.Ct. at 710-711. As Judge Markey has put it: When others in the field say “it won’t work,” that fact is strong evidence that whatever does work would have been nonobvious. Nickola v. Peterson, supra, 580 F.2d at 913. Here, by contrast, both Rodgers and Kessinger testified that if, prior to 1973, they had been presented with the problem they would have known, in light of the prior art, how to solve it. Of course the Court is aware of the inherent danger of the use of hindsight by such witnesses, but in reviewing all of the testimony on this issue, as well as the actual use of the core idea of “programmed delamination” at the United States Gypsum plant in the late 1960’s, the Court must conclude that, even taking it as a whole, as plaintiff urges, the Miller process was fatally obvious to a person reasonably skilled in the art. Such a person, in the Court’s view, “setting out to build a [multi-layer glass fiber mat] .. . could accomplish his mission without adding any ‘invention’ to the art.” Parker Sweeper Co. v. E. T. Rugg Co., 474 F.2d 950, 955 (6th Cir. 1973), cert, denied, 414 U.S. 829, 94 S.Ct. 56, 38 L.Ed.2d 63 (1973). Confronted with the problem of producing such a product it would have been obvious to do it by “programmed delamination,” i. e., the Miller process. Cf., Hanson v. Alpine Valley Ski Area, Inc., supra, 611 F.2d at 160. The Sixth Circuit’s opinion in Eisele v. St. Amour, 423 F.2d 135 (1970) is instructive here. In that case St. Amour patented a “hole location and concentricity gauge,” used to test the location and accuracy of holes drilled in metal automobile parts. Eisele, who had patented similar gauges, sought a declaratory judgment that St. Amour’s patent was invalid. Eisele’s gauges were difficult to use, and it was St. Amour’s improvement over Eisele’s prior devices which was the subject of the patent in suit. The district court found the patent valid and the Court of Appeals reversed, finding the device obvious. Speaking for the Court, Judge Edwards concluded: The relative simplicity of this change, the ease with which St. Amour accomplished it when requested to do so, and the promptness with which Eisele followed when St. Amour’s stationary indicator gauge proved successful all strongly suggest obviousness to one skilled in the art. As the courts have repeatedly noted, an ordinary mechanical improvement does not entitle the originator of it to a patent, no matter how useful or commercially successful it may prove to be. 423 F.2d at 137-38. Similar factors militate against a finding of “invention” here. The idea of altering a single mat process to produce multi-layer was fairly simple, patentee Miller accomplished it easily after he conceived the idea, and the Reichhold people followed with their own version promptly after the product of the Miller process appeared on the market. As with the patent in suit before the Supreme Court in Sakraida v. Ag. Pro., supra, Miller’s process is “doubtless a matter of great convenience, producing a desired result in a cheaper and faster way, and enjoying commercial Success” but likewise, however, it lacks the invention to make it patentable. In this Court’s view Miller “may have an improved product, but not an innovatively different one.” Cathodic Protection Service v. American Smelting & Refining Co., 594 F.2d 499, 511 (5th Cir. 1979), cert, denied, 444 U.S. 965, 100 S.Ct. 453, 62 L.Ed.2d 378. Secondary Considerations As noted earlier, Graham v. John Deere stated that such “secondary considerations” as long felt but unsolved needs, failure of *510 others to produce the patented item, and commercial success, may have relevance as indicia of obviousness or nonobviousness. 383 U.S. at 17, 86 S.Ct. at 693-694. These are considered below. Longfelt but Unsolved Need Both parties testified at trial that prior to 1973 when Nicofibers first began marketing its muti-layer product, there was no demand for such a product. As in Nickola v. Peterson, supra, 580 F.2d at 914, “there was no evidence that others had even tried to fill a need” for a multi-layer mat. The Court recognizes that the Miller process introduced certain economies of production into the glass fiber mat industry, but this factor alone cannot create patentability. “[A] cost savings is not a patentable difference. It is not a substitute for invention.” Elta Corp. v. Basie, Inc., supra, 599 F.2d at 757. Failure of Others Likewise the Court cannot find that others tried and failed to produce a multi-layer product. Nicofibers argues that following the advent of its Layer-Pak onto the market, Reichhold tried repeatedly without success to produce a similar product. Not until the Miller patent issued, claims the plaintiff, did Reichhold have the ability to make a multi-layer mat of its own. In this regard there was some evidence that Nicolet Industries briefly experimented with multi-layer mat in 1959 as an off-shoot of its research to determine the cause of delamination This research did not utilize the Miller process step of stopping the binder spray to produce mat separation. These experiments were at best only partially successful, and the process was never used commercially because, according to Robert Lovell, “there was no interest in it.” The Court cannot find that these brief and apparently isolated experiments amounted to a substantial effort to produce a multi-layer product. In addition, because these experiments were carried out prior to the general use of automatic spray equipment in the Modigliani process, the frequent malfunction of which greatly increased the occurrence and general knowledge in the industry of delamination, the Court declines to find that they have any substantial relevance in the consideration of the obviousness of the Miller process. Plaintiff also points to the fact that in the 1960’s one of the defendant’s customers routinely split defendant’s reinforcing mat in its manufacturing operations, and that in 1966 defendant recommended that a customer split defendant’s 10 mil mat to satisfy the customer’s need for 5 mil mat. Apparently plaintiff seeks to raise the inference that a better solution for these customers would have been a readily separable multi-layer product, and the further inference that, if the process for producing such a product were obvious, defendant would have found and used it. The Court cannot so find. A review of the testimony on these matters gives no indication that a multilayer product would have met the needs of the particular manufacturing processes. Indeed, there was evidence for example that a 5 mil mat is too thin to be produced reliably by the Modigliani process at all. In any event, there is no evidence that prior to 1973 Reichhold tried to produce a multilayer mat, much less that it tried to do so and failed. The patentee himself knew of no experiments, prior to 1973, by anyone seeking to make multi-layer mat. T-I-79. Plaintiff also claims that the unsuccessful work done in 1974 by defendant using talc as a release agent demonstrates that defendant was unsuccessfully searching for an answer to the multi-layer problem as late as that date. Although this may be probative of the fact that Reichhold was seeking methods to improve its process, it does not vitiate the evidence that Reichhold, when presented initially with the problem, knew how the Miller process worked; or in other words, that it was obvious. *511 Plaintiff also asserts that Reichhold’s manufacture of multiple mats using paper barriers between the layers constitutes an unsuccessful attempt by Reichhold to develop a multi-layer process. This is without foundation. Both Reichhold and Nicofibers still use the paper barrier process as an economical method to produce single layer mat for customers who require it. It is not necessarily a process for producing multilayer mat, and therefore does not, in the Court’s view, indicate a failure to develop such a process. Commercial Success Finally, turning to a consideration of the commercial success factor, there is no doubt that Nicofibers has developed a substantial market for its multi-layer product, while Reichhold sells virtually no multi-layer mat. Although this factor carries some weight, commercial success alone is not sufficient to establish that the product is the result of invention. Eltra Corp. v. Basic, Inc., supra, 599 F.2d at 756; Reynolds Metals Co. v. Acorn, supra, 548 F.2d at 162. The incisive analysis, and rejection, of the commercial success evidence produced in Nickola v. Peterson, supra, is applicable here: The primary reason for failure of the commercial success evidence in this case is its inadequacy.... An average of only 6,000 to 7,000 sales annually would not constitute a “rush to the invention” probative of nonobviousness. There was no evidence that a fortune had been waiting in the wings for him who first made the invention available to the public. It is that waiting fortune which implies that if the invention had been obvious it would surely have been earlier made. Further, there was no evidence of total demand, the extent to which that demand was met by the patented product, the extent to which the market abandoned earlier products in favor of the product patented, the extent to which advertising contributed or did not contribute to sales, and the like. In short, Nickola’s commercial success evidence did not rise to the level of an indicium of nonobviousness listed as potentially relevant in Graham v. John Deere . .. 580 F.2d at 914 (Markey, C. J., sitting by designation). Although not totally without weight, plaintiff’s commercial success evidence, in the Court’s view, suffers from the same shortcomings as that produced in Nickola. It is not enough, especially in view of other indicia of obviousness discussed above, to “tip the balance,” Eltra at 756, in Nicofibers’ favor. Conclusion The Court has carefully reviewed all of the evidence in this case, exercising particular caution to avoid reaching unwarranted conclusions through the use of hindsight. Upon such review, it is the opinion of the Court that in mid-1973 it would have been, and in fact was, obvious to a person of ordinary skill in the art, confronted with the problem of producing a multi-layer glass fiber mat, to have solved it by means of “programmed delamination,” that is, the Miller process. For this reason, and for all the reasons foregoing, the Court finds that the process fails to meet the statutory requirement of nonobviousness set forth in 35 U.S.C. § 103. Accordingly, the Court concludes and hereby DECLARES that the nine claims comprising the Miller patent are invalid. Because of this there is no need to reach the remaining issues of validity or infringement. Judgment will be entered for the defendant. So ORDERED. *512 Footnotes [1] The Miller patent consists of nine claims, as follows (see appendix for drawings): 1. Method of sequentially producing a plurality of glass fiber mats, comprising: a. continuously drawing glass filaments from a molten glass supply through a plurality of orifices; b. continuously winding the filaments about the periphery of a drum rotating about its axis; and c. intermittently applying a binder solution to the filaments being wound continuously on the drum, the binder solution being applied cyclically to the filaments at predetermined intervals during a first period of time sufficient in duration to form a first integral mat, the binder solution not being applied during a second period of time of predetermined duration such that an integral mat is not formed at a location, and the binder solution being applied cyclically to the filaments at predetermined intervals during a third period of time sufficient in duration to form an overlying second integral mat concentric with and separable from the first mat at said location where no binder solution is applied. 2. Method according to claim 1, wherein during the first and third periods, the binder solution is applied in cycles of 1 minute “on” and 1.1 minutes “off”. 3. Method according to claim 1, wherein' during the second period the binder solution is not applied for from about 5 to 12 minutes while the glass filaments are being wound continuously on the drum. 4. Method according to claim 1, wherein during the second period the binder solution is not applied for about 8.75 minutes while the glass filaments are being continuously wound on the drum. 5. In a method of producing glass fiber mats by continuously winding glass filaments about a rotating drum, the improvement comprising applying a binder solution at predetermined intervals to the filaments on the drum to form a plurality of concentric integral mats on the drum during periods of time wherein the binder solution is being applied, the mats being separable at a location where no binder solution is applied for a predeter *500 mined period of time sufficient in duration to prevent the formation of an integral mat at said location. 6. Method according to claim 5, wherein the binder solution is cyclically and uniformly applied to the filaments on the drum during said periods of time wherein the binder solution is applied. 7. Method according to claim 6, wherein the cycles are of one minute “on” and 1.1 minutes “off”. 8. Method according to claim 5, wherein said predetermined period of time sufficient to prevent the formation of an integral mat is of a duration of from about 5 to 12 minutes. 9. Method according to claim 5, wherein said predetermined period of time sufficient to prevent the formation of an integral mat is of a duration of about 8.75 minutes. The principal or independent claims of the Miller patent are claims 1 and 5. Claims 3, 6 and 8 are dependent upon claims 1 and 5 and relate to the 5 to 12 minute binder shut-off time parameter of the process. Claims 2 and 7 add to claims 7 and 5 that the binder spray is applied in cycles of one minute “on and 1.1 minutes” off. Finally, claims 4 and 9 modify claims 1 and 5 to state that the preferred binder shut-off time is 8.75 minutes. [2] The product of the Miller process-glass fiber mats-is identical to that of the Modigliani process. That product is not the subject of this dispute. Miller’s patent covers only his process for producing a plurality of concentric mats which can readily be separated. [3] In discussing the advantages of the Miller process, the Court recognizes that such advantages are generally only applicable to production intended for a customer who can utilize the multi-layer product. It is clear that some manufacturers prefer the single mats, and in fact both Nicofibers and Reichhold do a substantial business in the single, non-layered mats. [4] See, defendant’s Exhibits AW, AX, AY, AZ, BA, BB. [5] Kessinger has since attended a seminary and is now the pastor of a church in New York state. [6] Whether the phenomenon of delamination was considered by the Patent Office is somewhat unclear. It is true, as defendant points out, that the file wrapper makes no mention of delamination and that the patentee did not bring it before the examiner. On the other hand, in a preliminary review of Miller’s application, at which time several claims were rejected, the examiner commented as follows: The step of discontinuing the binder application reads on a clogged spray gun which intermittently fails to function. Defendant’s Exhibit AV. It may be inferred from this that the examiner was aware of delamination and that subsequent allowance of Miller’s amended claims indicates the Patent Office’s acceptance of the patent even with knowledge of that phenomenon. The defendant has not presented argument on this point. In any event, the Court finds that defendant has successfully met its burden of overcoming the statutory presumption accorded to plaintiff’s patent. [7] Defendant Reichhold introduced as evidence three prior art patents. They are as follows: U.S. Pat. No. Inventor Issue Date 2,081,060 P. Modigliani 5/18/37 2,546,230 P. Modigliani 3/27/51 3,342,658 R. L. Jackson, Jr. 9/19/67 In addition, the Court considers the relevant prior art to include the numerous other patents issued on the Modigliani process, cited in the Miller patent itself. The parties do not dispute that the various Modigliani patents are properly considered prior art.