Polysius Corp. v. Fuller Co. (709 F. Supp. 560, 1989)

OPINION

CAHN, District Judge.

In this complex patent litigation, plaintiffs seek declaratory and injunctive relief pertaining to United States Patent No. 4,357,287 which sets forth a method for reducing the size of brittle minerals. Plaintiffs also claim entitlement to counsel fees for the allegedly willful and deliberate conduct of the defendant. The defendant denies infringement and raises the defenses of patent invalidity, patent misuse, and fraud on the Patent Office. Defendant claims entitlement to counsel fees relating to the defense of the infringement count. The defendant also seeks money damages on its counterclaim for alleged antitrust and unfair competition violations.

The case was heard non-jury in February, March, and July of 1988. The parties, through counsel, have presented extensive closing arguments to the court and have filed post argument briefs as well as requests for findings of fact and conclusions of law. I make the following:

A. FINDINGS OF FACT

1. The plaintiffs are:

(a) Polysius Corp. (“Polysius”), a Georgia corporation with its principal place of business in Atlanta, Georgia;

(b) Dr. Klaus Schonert (“Schonert”), a citizen of West Germany and the owner of U.S. Patent No. 4,357,287 (the ’287 patent).

2. The defendant is Fuller Company (“Fuller”), a Delaware corporation with its principal place of business in Bethlehem, Pennsylvania.

3. The counterclaim defendant is Krupp Polysius A.G. (“Krupp Polysius”), a West German corporation. Polysius is a wholly-owned subsidiary of Krupp Polysius.

4. On February 17, 1978, Schonert filed a patent application in the United States Patent and Trademark Office for a “Method of Fine and Very Fine Comminution of Materials Having Brittle Behavior.”

5. (a) Prior to the filing of the United States application, Schonert granted a nonexclusive license to Krupp Polysius under a pending German patent application for the same method;

(b) Krupp Polysius thereafter embarked on a research and development program to design and manufacture equipment capable of carrying out the Schonert process. This program resulted in the development of a high pressure roll press marketed under the trade name “Polycom”;

(c) In 1979, a German firm, KlocknerHumboldt-Deutz (“KHD”) instituted an Opposition Proceeding to Schonert’s application in the German patent office;

(d) In 1985, Schonert, Krupp Polysius, and KHD entered into agreements which provided:

(i) KHD will withdraw its opposition to the issuance of the German patent;

(ii) Schonert will authorize Krupp Polysius to sub-license KHD to use the patented process;

(iii) Schonert will not grant further licenses without the prior approval of Krupp Polysius;

(iv) Krupp Polysius will not grant further sub-licenses without the prior approval of KHD;

(v) Krupp Polysius will not authorize Schonert to grant additional licenses without the prior approval of KHD;

(vi) Krupp Polysius and KHD will share expenses in regard to enforcing the patents for the Schonert process;

(e) KHD withdrew its opposition and the German patent issued.

6. On November 2,1982, the ’287 patent was issued to Schonert by the United States Patent Office.

7. The ’287 patent discloses and claims a process to comminute brittle material. The patent describes a process for fine and very fine comminution of brittle material in two steps. First, a bed of materials is compressed between non-yielding surfaces in a single pass at a pressure of at least 500kg/cm2. At such pressure the material is comminuted but is also agglomerated into flakes resembling pancakes. Second, the flakes are deagglomerated to release the fine material.

8. The major advantage of the '287 patent is significant energy savings (generally in the form of reduced electrical costs) for firms in the mineral, mining, and cement industries.

9. In the spring of 1985, Fuller sent a member of its engineering staff to meet with Schonert in Germany.

10. In June of 1985 Schonert met with the Fuller representative and explained his process in detail.

11. Thereafter, Fuller requested a license from Schonert under the '287 patent.

12. In July of 1985 Krupp Polysius refused to allow Schonert to license Fuller under the ’287 patent.

13. Thereupon, Fuller embarked on a program to develop equipment to compete with Polysius and others.

14. In June of 1985 Fuller’s Research and Development Department purchased a used Allis Chalmers compactor (“the A-C compactor”) and modified it as follows:

(a) a feed hopper was installed to create a bed of material above the rollers of the compactor;

(b) a 50 horsepower motor was incorporated rather than the 40 horsepower motor which was standard for this A-C compactor;

(c) the 8 inch roll width was reduced to 6 inches;

(d) devices were incorporated into the circuit to provide deagglomeration after the compaction step.

15. In August of 1985 Fuller used the rebuilt A-C compactor to carry out the Schonert process. Fuller used this equipment to analyze cement clinker samples obtained from prospective customers in an effort to sell high-pressure roll crusher equipment.

16. Fuller’s test procedure included stressing the brittle material provided by prospective customers by processing it through the counter-rotating cylindrical rolls of the A-C compactor in a single pass at pressures in excess of 500kg/cm2 which were sufficient to comminute the material. The agglomerates formed by this process were then disintegrated in subsequent steps.

17. In November of 1985 Fuller presented a seminar to the mineral and mining industries at its home office wherein the operation of the A-C compactor circuit was demonstrated and explained to those in attendance. A brochure was also distributed entitled “Fundamentals of the Clinker Preliminator.” This brochure was written by the engineer who visited Schonert.

18. Fuller prepared a videotape of the operation of the A-C compactor circuit. Upon advice of one of its patent lawyers, the audio portion of this videotape was erased.

19. Fuller designed and developed equipment for the express purpose of selling the equipment to carry out the Schonert process. Fuller’s equipment is known as the Fuller HRC.

B. DISCUSSION

The parties raise a plethora of issues all of which must be considered and discussed. The threshold issue, however, is the validity of the ’287 patent.

1. The ’287 Patent is Valid

a. The language of the ’287 patent complies with 35 U.S.C. § 112.

The first task in an adjudication of the issues of patent validity and infringement involves a consideration of the language of the patent in suit. Claim 1 of the ’287 patent reads as follows:

1. In a method for comminution of particles of brittle materials to obtain product of fine particles the maximum particle size of which is approximately in the range of 2 um-1000 um comprising:

(a) stressing a bed of bulk coarse material containing particles of the brittle material to be comminuted which are of a size larger than the fine particles to be obtained, between non-yielding hard surfaces to thereby obtain comminution of the coarse particles into finer fragments along with a agglomeration of finer fragments; and

(b) subsequently disintegrating the resulting agglomerates to obtain the product containing the fine fragments, the improvement comprising:

(c) carrying out said stressing of said bed of material by subjecting said bed to a pressure of at least 500kg/cm2 between said non-yielding surfaces; and

(d) carrying out said stressing in a single pass, whereby the total energy required [sic] comminute said particles will be substantially reduced.

The remaining claims alleged by plaintiffs to be infringed by Fuller are dependent upon Claim 1. Claim 2 discloses and claims the method of Claim 1, the difference being that Claim 2 requires pressure of at least 800kg/cm2. Claims 10 and 11 disclose and claim the method according to Claim 1, further specifying that the stressing referred to in Claim 1 be carried out using, respectively, a roller mill with cylindrical rotors and a roller briquetting press. Claims 15 and 16 disclose and claim the method according to Claim 1 with the deagglomeration being accomplished by, respectively, combined compression — sheer stressing and stressing in a ball mill. Claim 18 specifies that the method according to Claim 1 be carried out on such materials as cement clinker. Claim 19 discloses and claims a method according to Claim 18, specifying that the deagglomeration step take place in a ball mill. Claim 24 discloses and claims the method according to Claim 1, specifying that at least 30% of the fragments obtained after the deagglomeration step be in the size range of the final product.

35 U.S.C. § 112 provides in part:

The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, ... to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.

Section 112 requires the inventor to describe in the specification how the invention set forth in the claims actually works. The specification should explain what is new about the invention. This description and explanation must be sufficient to enable a person with skill in the field covered by the invention to replicate it without undue experimentation. This description must disclose the most advantageous method for carrying out the claimed invention.

Reference to Figures 4, 5a, and 5b in the ’287 patent and the text relating to them will provide sufficient information to decide whether Schonert has met the Section 112 requirements.

FIG. 4 shows how the compression force F, roller diameter D, roller axial length L, and maximum grinding pressure Pm which acts on the bed in the narrowest section of the nip are inter-related in accordance with the following equation:

F/LD = ka0Pm

wherein:

a0 = angle at which the stressing begins and

K = a constant dependent on the material behavior and having a value of about 0.2 in general.

The method according to the invention necessitates that the grinding pressure be selected high enough that a distinct formation of agglomerates or briquettes takes place. In general this requires the grinding pressure Pm to be increased to more than 500 kg/cm 2 or 7000 psi. The compression force of the rollers per meter of roller length (F/L) results, in accordance with the above equation, as being proportional to the roller diameter D and the chosen angle a0. For the value F/L = 200000 kg/m given above and for D = 100 cm, a0 = 0.1 (corresponding approximately to 6°), and k = 0.2, a grinding pressure of Pm = 1000 kg/cm2 or 14000 psi results.

Tests have shown that the method according to the invention permits considerable savings of energy. In the production of normal Portland cement PZ275 the comminution requires only from 10 to 20 kWh/t instead of 25 to 35 kWh/t. The new method can be carried out in a plant of the type illustrated by the block or flow diagram of FIG. 5a. It consists of a mill 16 called GWM. Mill 16 is a roller mill for grinding a bed of bulk material, for instance, a roller mill having two cylindrical rollers driven in a counter rotating sense, such as shown in FIGS. 2 and 4, or another type of roller mill; a ball mill KM 17, and an air classifier K1 18 having its separation limit xtr at 60 um. The pre-crushed material, all the particles xa of which are smaller than 2.5 mm, is supplied to mill 16 at a feed rate M and the recycled coarse material from the air classifier K1 18 is supplied to the mill 16 at a feed rate Mg. The stressing between the rollers effects comminution and briquetting. The resulting product is compressed to flake-shaped briquettes which are disagglomerated or disintegrated in the ball mill 17 connected downstream. The product leaving the ball mill contains approximately 40% of particles which are smaller than 60 um. The air classifier K1 18 separates the major part thereof into an end product which leaves the mill-classifier circuit at a product rate M. The recycled coarse material is mixed with the pre-crushed feed material and again supplied to roller mill 16. The total flow rate M* through the mill is divided by the air classifier K118 into two fractions, the fine fraction leaving the circuit at the product rate M = p’M*, the coarse fraction rate being recycled at a flow rate (l-p’)M*. If the product rate is assumed, for example, to be M = 100 t/h and p’ = 33%, the flow rate through the mill and the air classifier will be M* = 300 t/h. The energy required by the roller mill 16 for the bed of bulk material amounts to 3.1 kWh/t and that of the ball mill 17 amounts to 1.4 kWh/t. The sum of both is 4.5 kWh/t. The specific energy consumption of the circuit based on the production rate is calculated to amount to approximately 13.6 kWh/t. If, furthermore, 20% of the gross energy input is considered to the [sic] lost in the engine transmission, and machines, the resulting specific energy consumption will be about 17 kWh/t.

FIG. 5b shows a block flow diagram of plant comprising two roller mills for grinding a bed of bulk material, GWM1 19 and GWM2 20, one ball mill 21, and two air classifiers Kll 22 and K12 23. Subdividing the stressing of the bed of bulk material into two steps establishes an advantage from the point of view of the process technique and may also be favorable with respect to the overall economy.

When applying the novel method, the finest comminution of limestone to a product of which 100% is smaller than 10 um requires compression of about 1,500 kg/cm2 (2,000 psi) and only about 10 kWh/t. Optimum operating conditions can be found for every task. For example, the pulverization of cement is most economical at grinding pressure between 1000 kg/cm2 or 15,000 psi and 2,500 kg/cm2 or 35,000 psi.

Fuller maintains that the claims and specification do not meet the requirements of Section 112. Fuller claims that all of the product obtained following the deagglomeration step (KM-17 in Fig. 5a and KM-21 in Fig. 5b) must be in the size range of 2 microns to 1000 microns. According to Fuller, if some of the product following the deagglomeration step exceeds 1000 microns in size, there would be no infringement of the '287 patent. As a corollary to this argument, Fuller urges that if the ’287 patent is interpreted so that some of the product following the deagglomeration step may exceed 1000 microns in size, then one would be unable to ascertain at what point he would be infringing the patent. With this approach, Fuller hopes to place the plaintiffs on the horns of a dilemma. If the 2 to 1000 micron range in Claim 1 is absolute, then Fuller could avoid the patent by arranging its operation so that some of that product exceeds 1000 microns in size. On the other hand, if Claim 1 is interpreted so that some of the product following the deagglomeration step may exceed 1000 microns in size, then the claims are too indefinite and violate Section 112.

I hold that the reference in Claim 1 is to an approximate range of 2 to 1000 microns.

Fuller emphasizes in Claim 1 of the ’287 patent that the stressing is carried out in a single pass. Fuller urges that the 2 to 1000 micron range set forth in Claim 1 must be achieved in the first pass through the high-pressure roll press without recirculating particles larger than 1000 microns for further comminution.

There are several serious problems with Fuller’s interpretation. In the first place, Claim 1 qualifies the 2 to 1000 micron range by using the word “approximately.” Second, Fuller’s interpretation is totally impractical. The precise range of fineness of the product following the deagglomeration step has little or nothing to do with the novelty of the ’287 patent. The '287 patent, contrary to prior mineral engineering practice, teaches that high-pressure stressing of a bed of materials between non-yielding surfaces is an energy saving comminution technique, even if agglomerates are formed. The prior practice was to avoid the formation of agglomerates because the sole purpose of comminution is to achieve a reduction in size and the formation of agglomerates, it was thought, defeats that purpose. See Record at 8.131.

Schonert discovered that the pancake-like agglomerates formed when brittle material is stressed in accordance with the ’287 patent can easily and efficiently be deagglomerated. Fuller’s emphasis on the precise range of fineness for the product following deagglomeration is misplaced. It makes no difference if some of the product following deagglomeration is in a range of fineness that exceeds 2 to 1000 microns. Fuller is being entirely too literal in insisting that the patent be read so that no further comminution may take place following the deagglomeration step. In commercial grinding applications it is customary for oversize fragments to be recirculated through the comminution process. In fact, Fig. 5a and Fig. 5b illustrate classification equipment which will recirculate oversize material through the high-pressure roll press. In Fuller’s view, if any of this material exceeds 1000 microns in size, there could be no infringement of the ’287 patent. Fuller argues that recirculation of product in excess of 1000 microns would violate the single pass concept of Claim 1.

I see no reason to find non-infringement in the situation where an operator utilizing the concepts of the '287 patent adjusts his equipment so that some product following deagglomeration is more than 1000 microns in size and is subsequently classified and recirculated to reduce it to a smaller final product size. Instead, I think the correct interpretation of the ’287 patent is that the 2 to 1000 micron range is merely approximate. Otherwise, it would be ridiculously easy to circumvent the patent by adjusting the operation to produce a small amount of product following deagglomeration to be in excess of 1000 microns in size.

Puller has hypothecated in its cross-examination of plaintiffs patent law expert and in its post-argument briefing certain situations where, if plaintiff’s interpretation is adopted, a potential infringer would have difficulty in determining at what point he would infringe the ’287 patent. Fuller asks the rhetorical question: “Would a comminution process be within Claim 1 if the amount of fines in the product exceeds the amount of fines in the feed by 10 percent and the purported energy savings is in the order of 5 percent?” The answer is obviously “no”, unless the process utilizes the stressing of a bed of brittle material between unyielding hard surfaces at a high pressure with subsequent deagglomeration and with almost all of the comminution occurring in the first pass. This hypothetical inquiry as well as the other rhetorical questions asked by Fuller in its post argument briefing illustrate the lack of practicality with Fuller’s position. What is important here is the concept that high pressure grinding of a bed of materials will result in substantial energy savings and a product following deagglomeration with most of the particles being of a size less than 1000 microns.

“Further, claims should be so construed, if possible, as to sustain their validity.” ACS Hosp. Systems v. Montefiore Hosp., 732 F.2d 1572, 1577 (Fed.Cir.1984). Fuller’s attempt to avoid the ’287 patent on the ground that the patent applies only if all of the product following the deagglomeration step is 1000 microns in size or smaller is unpersuasive and will be rejected.

Fuller maintains that the best mode for carrying out the Schonert invention is not disclosed in the ’287 patent. Reference to the patent itself, as well as reference to the extensive evidence of the utilization of the Schonert process, establishes that the best mode appears to be the use of hydraulically operated counterrotating roll presses with subsequent deagglomeration in a ball mill. Schonert has fully and candidly disclosed this mode. Also, there is no evidence of concealment by Schonert of a better mode not disclosed in the '287 patent. See Hybritech, Inc. v. Monoclonal Antibodies, Inc., 802 F.2d 1367, 1384 (Fed.Cir. 1986), cert. denied, 480 U.S. 947, 107 S.Ct. 1606, 94 L.Ed.2d 792 (1987).

b. The prosecution history of the ’287 patent.

The discussion in this section will be limited to a review of the prosecution history related to Fuller’s contention that all of the product size following deagglomeration must be in the range of 2 to 1000 microns.

In an amendment filed with the United States Patent and Trademark office in late 1981, Schonert, after thanking the examiner for granting his United States and German attorneys an interview, stated:

As discussed with the examiner at the interview, the claims have been completely rewritten. Claim 47, (now Claim 1) the independent claim replacing Claim 22, has been written in Japson [sic] format. In addition to specifically pointing out the steps which comprise the present invention, the claim has been limited to the range of 2 um to 1000 um. This avoids any indefiniteness by what was previously defined as fine or very fine.

In the preceding section I held that not all of the product following the deagglomeration step must be in the 2 to 1000 micron range. The prosecution history does not undermine this holding. My interpretation of the range in Claim 1 is consistent with dependent Claim 24. Claim 24 states:

24. The method according to Claim 1 comprising selecting the compression force for carrying out said stressing to be of sufficient magnitude so as to result in at least 30 percent of the fragments obtained having the desired degree of fineness of the final product.

Claim 24 refers to applying a compression force of sufficient magnitude so that 30 percent of the product following deagglomeration would be as small or smaller than the desired degree of fineness for the end product. In Fig. 5a of the patent, for a process to meet Claim 24, 30 percent of the fragments following step 17, the deagglomeration step, must be smaller than 60 microns.

Claim 24 is dependent upon Claim 1. Both sides use the dependency of Claim 24 to support their respective positions on the maximum size of the product following deagglomeration. Plaintiffs urge that if one assumes a desired final product size following classification of 1000 microns, then to meet Claim 24, 30 percent of the product following deagglomeration must be at least that small. But, conversely, 70 percent of the product following deagglomeration would then be over 1000 microns in size. Consequently, plaintiffs maintain that the 2 to 1000 micron range in Claim 1 is approximate and not absolute because otherwise, Claim 1 could become more narrow than Claim 24, which by definition is not possible.

Fuller buttresses its position that no product size following deagglomeration may exceed 1000 microns by claiming that Schonert agreed to the 2 to 1000 micron range to comply with the Patent Examiner’s desire to quantify what Schonert meant by “fine or very fine”. According to Fuller, it is more logical to assume a desired final product size of 75 microns and that Claim 24 would then require 30 percent of the product following deagglomeration to be 75 microns or less with the remaining 70 percent of that product being in the range of 75 to 1000 microns, but in no event any larger. While there is some merit to Fuller’s position, it runs contrary to common sense and industrial practices. This court will not strain to interpret patent claims so that a patent holder is denied the fruits of his labor on technicalities. As stated, infra, it would be a travesty to allow Fuller or others to avoid the ’287 patent by adjusting the operation to produce product at the deagglomeration stage larger than 1000 microns in size,

c. The ’287 patent is not obvious under 35 U.S.C. § 103.

In determining whether the teaching of a patent is obvious, I should make the factual determinations required by Specialty Composites v. Cabot Corp., 845 F.2d 981, 989 (Fed.Cir.1988) which are “(1) the scope and content of the prior art; (2) the differences between the prior art and the claims at issue; (3) the level of ordinary skill in the art at the time when the invention was made; and (4) objective evidence of nonobviousness.”

The primary dispute between Fuller and Polysius is whether or not Taggart contains data that makes the teaching of the ’287 patent obvious. Consequently, I will review Fuller’s contention that Taggart discloses the teaching of the ’287 patent. I will then turn my attention to whether or not other prior art renders the ’287 patent obvious. On the issue of obviousness, Fuller has the burden of proof.

Taggart contains over 1000 pages relating to comminution technology. Fuller concentrates on two particular references in Taggart which, in its view, suggest that roll presses, operating at high pressures, have been used to comminute minerals which are then further comminuted (deagglomerated) in a ball mill. Taggart, in considering preliminary reduction, at page 6-45 states:

On Clinker (Table 20) there was an increase of 30 percent in capacity of a mill consuming 400 hp., or a saving of 7.2 hp-hr. per ton, by reason of a preliminary reduction through the range from <% in. to <6-m. Such reduction can be done for half or less of this cost in a fine crusher. The difference was even more striking in the case of limestone (Table 21). Table 18, item 213, shows a saving of slightly more than 50 percent in power consumption with a small tumbling mill for reduction in feed size from <2 in. to <20-m.

Fuller urges that this reference in Taggart is proof positive that fine crushing prior to use of a ball mill was a known method to achieve energy savings. Fuller’s contention, however, is in error because it has not carefully read Taggart. The next sentence in Taggart states:

A large part of this saving would have been made if the feed-size reduction had been to xk inch only (see Table 21), and the relatively costly preliminary reduction from