ARCONIC CORPORATION v. NOVELIS INC.

Opinion

IN THE UNITED STATES DISTRICT COURT FOR THE WESTERN DISTRICT OF PENNSYLVANIA

ARCONIC CORPORATION AND ) CIVIL ACTION NO. 17-1434 HOWMET AEROSPACE INC., ) ) JUDGE JOY FLOWERS CONTI Plaintiffs and Counterclaim Defendants, ) ) ) ) v. ) ) NOVELIS INC. and NOVELIS CORP, ) ) Defendants and Counterclaim Plaintiffs. )

MEMORANDUM OPINION I. Introduction Pending before the court are numerous Daubert motions filed by Arconic Corporation and Howmet Aerospace, Inc. (collectively “Arconic”) and Novelis Inc. and Novelis Corporation (collectively “Novelis”). In this opinion, the court will address the motions related to Arconic’s breach of contract claims with respect to the seven items of alleged confidential information (the “7 CI”) identified by the special master in report and recommendation (“R&R”) #33 (ECF No. 510 at 57-61). Specifically, this opinion will address the Daubert motions with respect to: (1) Arconic’s damages expert, James W. Bergman (“Bergman”) (ECF No. 839); (2) Arconic’s technical expert, Dr. Frank Ernst (“Ernst”) (ECF No. 849); (3) Arconic’s rebuttal expert on nondisclosure agreements (“NDAs”), Russell Stalters (“Stalters”) (ECF No. 852); and (4) Novelis’ technical expert, Dr. Timothy J. Eden (“Eden”) (ECF Nos. 861, 864 – under seal). The remaining Daubert motions were referred to the special master for R&Rs. The court held two days of oral argument on June 28 and June 29, 2022. The special master was present and able to participate. (Transcripts, ECF Nos. 935, 940, under seal). The court directed post-hearing briefing. The parties filed their respective positions (ECF Nos. 934, 945) and the motions are ripe for disposition. Although many of the related filings are sealed,

the court concludes that this opinion will not be sealed.

II. Factual and Procedural Background Arconic and Novelis are competitors in the aluminum industry. Ford Motor Company (“Ford”) decided to make its popular F-150 pickup truck with aluminum, starting with the 2015 model year. Arconic’s A951 pretreatment process (the “A951 process”) was selected for exclusive use in the Ford F-150 project. Ford was unwilling to be dependent upon a sole supplier. As a condition of selection, therefore, Ford forced Arconic to license its A951 technology to Novelis. Arconic’s 7 CI claims sound in breach of contract. In 2011, Arconic and Novelis entered

into a Confidentiality, Nondisclosure and Limited Use agreement (“NDA”) and in 2012 they executed a technology access and license agreement (“License”) (ECF Nos. 177-1, 177-2). Count II of Arconic’s operative second amended complaint involves the License and count IV involves an NDA.1 Count II will be subject to a bench trial because the License contains a jury waiver provision. Whether count IV will be tried to a jury depends on which NDA is at issue.2

1 There are two NDAs potentially at issue. At the oral argument, Novelis notified the court that it intends to seek summary judgment because one NDA expired before Novelis filed the patent application and the other NDA does not address confidential information. (6/28/22 Tr., ECF No. 935 at 6). For the purpose of this opinion, the court will assume a valid NDA exists. 2 Arconic’s request for declaratory judgment at Count VII (ion exchange) is required to be resolved by the court, not a jury. There are no Daubert motions pending with respect to the competing expert opinions involving ion exchange. The License provided, among other things, that Novelis could use processes that were part of general industry practices, retained its own technology and know-how, and owned improvements it developed independent of Arconic. License §§ 2.2.1, 2.2.2, 2.2.3. Arconic’s disclosures to Novelis took place in the context of three patents related to the A951 process,

which by reason of the public nature of patents involved disclosures in the public domain. The breach alleged by Arconic occurred in a one-time event six years ago. On November 3, 2016, Novelis filed patent application No. 0319440 (the “’440 patent”) with respect to improvements it made to the A951 process. Arconic filed this lawsuit in 2017, alleging that Novelis disclosed Arconic’s trade secrets and confidential information in the ‘440 patent application. No threat of ongoing or future misconduct by Novelis was raised by the pending claims. Arconic’s claims are not based on the chemical formulation (“A951”) used in the A951 process – the specific chemical formulation was never disclosed to Novelis. (ECF No. 269, Ex. K at 16, filed under seal). Because the alleged disclosures in the ‘440 patent application involved already-public

information from Arconic’s prior patents and Novelis’ own improvements, the court explained (repeatedly) that before allowing Arconic to undertake extensive discovery into its competitor’s technology, it was incumbent upon Arconic to first articulate, with specificity, what it contends are its own trade secrets and confidential information, as opposed to general industry practices or Novelis’ authorized improvements to the A951 process. As this court has recounted at length, Arconic never made an adequate disclosure. Ultimately, the court granted summary judgment in favor of Novelis on the entirety of Arconic’s trade secret claims and on the vast majority (281 of 288) of Arconic’s confidential information claims. The confidential information claims in counts II and IV were permitted to proceed only on the narrow basis identified by the special master in R&R #33 (i.e., with respect to the 7 CI) (ECF Nos. 622, 623). Arconic admits that it suffered no actual damages caused by Novelis’ disclosure of the 7 CI in the ‘440 patent application in 2016. There is no evidence of lost royalties or lost sales by

Arconic. Bergman Deposition (ECF No. 841-3) at 169 (“I haven’t seen any evidence of lost profits”); Arconic Rule 30(b)(6) Deposition through designated representative David Coates (“Coates Deposition”) (ECF No. 841-5 at 339 (“At this point in time, we have not positively identified any lost profits that we can correlate to that disclosure. . . . At this point, we do not have any direct evidence that we have lost sales due to the disclosure of the seven CI”). Arconic distributes A951 through an exclusive supplier, Chemetall GmbH (“Chemetall”). Arconic receives royalties on all sales involving Chemetall. Arconic receives royalties on all purchases of A951 by Novelis, who obtains it through Chemetall. Coates Deposition at 341 (“Arconic is not aware of any aluminum sheet supplier that pretreats aluminum with A951 that would have procured or obtained that pretreatment from anybody other than Chemetall.”).

Chemetall, Novelis and Ford all benefited enormously from the F-150 project. (Mark A. Israel (“Israel”) Report at 23). Arconic recognizes that it benefited from the F-150 project as well. (6/29/22 Tr., ECF No. 940 at 31) (“Of course, Arconic benefits too.”). The A951 process has been running commercially for the past six years at Arconic, Novelis and Material Sciences Corporation (“MSC”), a tolling manufacturer. Ford has purchased millions of pounds of A951- treated aluminum. (Bergman Report, Exh. 9). III. General principles of law A. Daubert In Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 594-95 (1993), the Supreme Court explained that the burden is on the party offering expert opinion to show: (1)

qualifications; (2) methodology/reliability; and (3) fit. See Fed. R. Evid. 702. The disputes in this case primarily involve “fit.” For expert testimony to meet the “fit” requirement, it must “assist the trier of fact to understand the evidence or to determine a fact in issue.” Fed. R. Evid. 702. As explained in Jaasma v. Shell Oil Co., 412 F.3d 501, 513 (3d Cir. 2005), there must be a “relevant connection between that methodology and the facts of the case.” To “fit,” the expert opinion must be relevant. “Expert testimony which does not relate to any issue in the case is not relevant and, ergo, non-helpful.” Daubert, 509 U.S. at 591 (citations and internal quotation marks omitted). The proffered expert testimony “must in fact assist the jury, by providing it with relevant information, necessary to a reasoned decision of the case.” Gallatin Fuels, Inc. v. Westchester

Fire Ins. Co., 410 F. Supp.2d 417, 420 (W.D. Pa. 2006). Courts will reject expert opinions about future damages if those opinions are not accompanied by a sufficient factual foundation. Elcock v. Kmart Corp., 233 F.3d 734, 755-56 (3d Cir. 2000) (concluding that district court abused its discretion by admitting a poorly supported damages opinion into evidence). B. Breach of contract elements As noted above, the expert opinions addressed in this opinion relate to Arconic’s breach of contract claims involving the 7 CI. The parties agree that Pennsylvania law applies. “Under Pennsylvania law, a breach of contract claim has three elements: (1) the existence of a contract and its essential terms, (2) a breach of a duty or obligation under the contract, and (3) harm resulting from the breach.” Lemons v. Meguerian, No. CV 21-1737, 2022 WL 1289128, at *6 (E.D. Pa. Apr. 29, 2022) (citing CoreStates Bank, N.A. v. Cutillo, 723 A.2d 1053, 1058 (Pa. Super. Ct. 1999)). The causation element (i.e., that harm resulted from the breach) must be proven before

the court reaches any dispute about the measure of damages. As explained in Brader v. Allegheny General Hospital, 64 F.3d 869 (3d Cir. 1995): In order to state a claim for damages arising from a breach of contract, a plaintiff must also plead damages resulting from the alleged breach. See General State Auth. v. Coleman Cable & Wire Co., 27 Pa.Cmwlth. 385, 365 A.2d 1347, 1349 (1976). This is a natural extension of the general rule that damages for breach of contract are not recoverable unless there is a “causal relationship between the breach and the loss.” See Robinson Protective Alarm Co. v. Bolger & Picker, 512 Pa. 116, 516 A.2d 299, 303 n. 9 (1986).

Id. at 878; see Best Med. Int'l, Inc. v. Buchanan Ingersoll & Rooney PC, No. CV 20-1077, 2021 WL 2954103, at *6 (W.D. Pa. Mar. 15, 2021), report and recommendation adopted sub nom. Best Med. Int'l, Inc. v. Buchanan Ingersoll & Rooney P.C., No. CV 20-1077, 2021 WL 2953656 (W.D. Pa. July 14, 2021) (“it is well-established that damages for breach of contract are not recoverable unless there is a causal relationship between the breach and the loss.”) (quoting Brader). In this case, to succeed on a breach of contract claim Arconic must establish that it suffered harm resulting from Novelis’ disclosure of the 7 CI in the 2016 patent application. See Hanwha Azdel, Inc. v. C & D Zodiac, Inc., No. 6:12-CV-00023, 2014 WL 2452892 at *14 (W.D. Va. June 2, 2014), aff'd in part, rev'd in part and remanded, 617 F. App'x 227 (4th Cir. 2015) (“the question is whether the alleged disclosures caused Plaintiff to lose any Aero-lite sales the company might otherwise have had, or whether the disclosures caused Plaintiff any other type of harm.”). In Hanwha, which involved a similar dispute about whether a disclosure of confidential information resulted in damages, the court explained: “Causation is an essential element of proof of liability in a breach of contract case, and it is an element that must be proven before the measure or amount of damages in any amount can even be addressed.” Id. See Hand v. Am. Bd. Of Surgery, Inc., No. CIV.A.01-2172, 2002 WL 227174, at *6 (E.D. Pa. Feb. 14, 2002), aff'd, 53

F. App'x 663 (3d Cir. 2002) (failure to establish a causal connection between the breach and damages neglects an essential element of a breach of contract claim); Time Peace Equine, Inc. v. Santander Bank, N.A., No. 4:17-CV-01561, 2018 WL 461369, at *3 (M.D. Pa. Jan. 18, 2018) (dismissing breach of contract claim where plaintiffs did not allege “any damages resulting from the breach, as is required under Pennsylvania law.”).

C. Kinds of damages for breach of contract Under Pennsylvania law, there are three distinct theories of damages for breach of contract: (1) expectation damages; (2) reliance damages; and (3) restitution damages. ATACS Corp. v. Trans World Commc'ns, Inc., 155 F.3d 659, 669 (3d Cir. 1998). The preferred theory is

expectation damages, which seeks to protect the “benefit of the bargain.” Id. If it is difficult to determine expectation damages (for example, if lost profits cannot be measured with certainty), Pennsylvania contract law provides for reliance or restitution damages, which seek to return the parties to the position they occupied before the contract was executed. Id. Expectation damages are “designed to place the aggrieved [party] in as good a position as would have occurred had the contract been performed.” Id. “[E]xpectation damages are measured by ‘the losses caused and gains prevented by defendant's breach, to the extent that [they] are in excess of any savings made possible by nonperformance.’” Id. (quoting American Air Filter Co. v. McNichol, 527 F.2d 1297, 1299 (3d Cir.1975) (citations omitted). Reliance damages protect an injured party's reliance interest “by seeking to achieve the position that it would have obtained had the contract never been made, usually through the recovery of expenditures actually made in performance or in anticipation of performance.” Id. In this case, Arconic seeks to recoup R&D costs it incurred long before it entered the contracts at

issue. Restitution damages “require the party in breach to disgorge the benefit received by returning it to the party who conferred it.” Id. Pennsylvania courts look to traditional principles of equity, such as unjust enrichment or forfeiture, in considering the propriety of restitution damages. Id. In Hansen Bancorp, Inc. v. United States, 367 F.3d 1297, 1315 (Fed. Cir. 2004), the court explained that “if the value of the benefit conferred upon the breaching party is not ascertainable through some usable measure of damages, restitution may not be possible.” In Fishkin v. Susquehanna Partners, G.P., 340 F. App'x 110 (3d Cir. 2009) (rejecting restitution damages where party failed to prove the value of the benefit it conferred), the Third Circuit Court of Appeals explained that restitution damages are not measured by the breaching party’s

profits, but by the value conferred by the aggrieved party: the ATACS Court's recognition of the possible remedy of restitution damages as measured by “ ‘the fair value of [the subcontractor's contribution]’ ” corresponds to Restatement (Second) of Contracts § 344's characterization of a party's restitutionary interest as the “interest in having restored ... any benefit that [it] has conferred on the other party.” See Trosky, 652 A.2d at 817 (discussing § 344).

Id. at 117. In CSS, Inc. v. Herrington, 306 F. Supp. 3d 857, 882 (S.D. W.Va. 2018), the court rejected an unjust enrichment claim where the plaintiff failed to identify the alleged benefit conferred upon the defendants in connection with their unlawful use and disclosure of confidential information. In addition, restitution damages are not appropriate where the property cannot feasibly be returned to the original party. As explained in Hansen: Finally, section 384 of the Restatement precludes restitution when the non- breaching party cannot return “any interest in property that he has received in exchange in substantially as good condition as when it was received by him....” Thus, if the non-breaching party has used, destroyed or substantially altered in character property received from the breaching party so that return is not feasible, restitution is generally not available. Restatement (Second) of Contracts 384 cmt. a.

Id. Restitution is not an appropriate remedy for disclosure of confidential information – once the 7 CI were publicly disclosed in a patent application, that information cannot be returned to the exclusive use of Arconic, i.e., the court cannot “unring the bell.” Returning the 7 CI to Arconic would frustrate the underlying purpose of the contracts because Ford required Arconic to share that information with Novelis as a condition for selecting A951. In other words, the return of the 7 CI to Arconic would not be feasible or appropriate.

D. Damages cannot be speculative Damages cannot be speculative. “As a general rule, damages are not recoverable if they are too speculative, vague or contingent and are not recoverable for loss beyond an amount that the evidence permits to be established with reasonable certainty.” Spang & Co. v. U.S. Steel Corp., 545 A.2d 861, 866 (Pa. 1988). Under Pennsylvania law, the evidence must afford a reasonable basis to support an award of damages; a verdict cannot be based upon mere speculation or guesswork. Helpin v. Trustees of Univ. of Pa., 10 A.3d 267, 270 (Pa. 2010). “Damages for lost profits, like other contract damages, may not be awarded when the evidence leaves the trier of fact without any guideposts except his or her own speculation. Sufficient evidence must be introduced to permit a reasonably certain estimate of the amount of anticipated profits lost due to the breach.” Merion Spring Co. v. Muelles Hnos. Garcia Torres, S.A., 462 A.2d 686, 695 (Pa. Super. Ct. 1983). Relief cannot be “too speculative, vague or contingent upon some unknown factor.” ATACS, 155 F.3d at 669. The law makes a distinction, however, between the existence of damages and the amount of damages. See Wachovia Bank N.A. v. Feretti, 935 A.2d 565, 572 (Pa. Super. Ct. 2007)

(“damages are speculative only if the uncertainty concerns the fact of damages rather than the amount.”). Pennsylvania law allows for some uncertainty in calculating the amount of damages for breach of contract. ATACS, 155 F.3d at 670; see Kivett v. Neolpharma, Inc., No. 2:20-CV- 00664, 2022 WL 1185885, at *6 (E.D. Pa. Apr. 21, 2022) (refusing to award damages where there was no evidence regarding future sales, but noting “[b]reach of contract damages cannot be based on a guess or speculation, but courts can fairly estimate them from the evidence, even if there is not mathematical certainty.”).

E. No windfall As Arconic recognized in its post-hearing brief (ECF No. 934 at 1-4), Pennsylvania

contract law damages awards do not entitle a party to receive a windfall. “The measure of damages for breach of contract is compensation for the loss sustained. The aggrieved party can recover nothing more than will compensate him.” Helpin, 10 A.3d at 270 (emphasis in original). The objective of expectation damages is to place the non-breaching party “as nearly as possible in the same position [it] would have occupied had there been no breach.” Id. (citation omitted). The “no windfall” principle is equally applicable to reliance and restitution damages awards. The purpose of reliance damages is “to place injured parties in the same position as they were prior to the execution of the contract, not to bestow a windfall.” Spring Creek Expl. & Prod. Co., LLC v. Hess Bakken Inv., II, LLC, 887 F.3d 1003, 1026–27 (10th Cir. 2018), as revised (Apr. 13, 2018) (applying Pennsylvania law) (noting absence of any decision “in which a plaintiff was allowed to pursue—let alone recover—reliance damages in excess of ascertainable expectation damages.”). In Hansen, the court explained: While an award of restitution should seek to return the non-breaching party to as good of a position as it would have been in if the contract had never been entered into, “the non-breaching party should not be placed in a better position through the award of damages than if there had been no breach.” Bluebonnet Sav. Bank, F.S.B. v. United States, 339 F.3d 1341, 1345 (Fed. Cir. 2003). Courts should avoid bestowing an “unfair windfall” on the plaintiff by compensating him or her above and beyond the losses suffered under the breached agreement.

Hansen, 367 F.3d at 1315. In Glick v. Campagna, 613 F.2d 31 (3d Cir. 1979), the Third Circuit Court of Appeals explained: The purpose of rescission and restitution is to restore the plaintiff to where he would have been had the sale not taken place. To place him in a better position is beyond the scope of the remedy and amounts to punishment of the defendant.

Id. at 37.

F. Nominal damages Under Pennsylvania law, the remedy for breach of contract where there are no actual damages is not the award of speculative or windfall damages. Instead, Pennsylvania law provides for recovery of nominal damages. See Wolfe v. Allstate Prop. & Cas. Ins. Co., 709 F.3d 487, 497–98 (3d Cir. 2015) (“Under Pennsylvania law, if a party is able to prove breach of contract but can show no damages flowing from the breach, the party is entitled to recover nominal damages.”); Zeno v. Ford Motor Co., 480 F. Supp. 2d 825, 837 (W.D. Pa. 2007) (need to prove causation does not apply to a claim for nominal damages); Garcia v. Vertical Screen, No. CV 19-3184, 2022 WL 829767, at *12 (E.D. Pa. Mar. 21, 2022) (“Because nominal damages are available for breach of contract under Pennsylvania law, courts have refused to grant summary judgment on a breach of contract claim solely because a party has not established damages.”).

IV. Bergman damages opinions

With that background, the court turns to Novelis’ motion to exclude Bergman’s opinions with respect to damages resulting from disclosure of the 7 CI (ECF No. 839).3 As noted, there are no actual damages claims in this case; e.g., there are no lost sales or lost royalties by Arconic. Future damages, however, are asserted. Bergman, Arconic’s expert, offers two theories of future damages: (1) a running royalty on future sales; and (2) recoupment of research and development (“R&D”) costs. Bergman does not offer an opinion on royalty damages. Instead, he suggests that the court should merely determine a royalty rate, which would then be applied in a future proceeding to unidentified future sales. With respect to recoupment of R&D costs,4 Bergman offers two contradictory opinions: (1) all $4.9M in development costs incurred from 1992 through 2012 are recoverable because they cannot be apportioned (Bergman Report ¶ 152)5; and

(2) an apportionment methodology – i.e., 5.67% of the total A951 experiments involved the 7 CI, so Arconic can recoup 5.67% of the $4.9M (roughly $278K). Bergman Report ¶ 160.

3 To be clear, Novelis is not seeking to exclude Bergman’s rebuttal to Julie Davis’ expert opinions. 4 The court understands Arconic to be asserting a “restitution damages” theory. To the extent Arconic seeks to characterize these as “reliance damages,” (see ECF No. 934 at 8 n.5), they are clearly not recoverable. There is no evidence that Arconic incurred these costs (many years earlier) in reliance on Novelis’ entry into the contracts at issue. Moreover, trying to put Arconic back in the position it would have obtained had the contracts never been made, see ATACS, 155 F.3d at 669, would require recognition of the millions of dollars Arconic has made from the F-150 program. 5 Bergman’s opinion that Arconic can recover all $4.9 M of its development costs is irreconcilable with the court’s ruling that Arconic can only recover for the 7 CI, and not the other 281 claimed CI, see ECF No. 623. Because it is contrary to the law of the case, that opinion will not be admissible under any circumstance. Bergman opines that Arconic is entitled to future royalties based on the value of the A951 process. (Bergman Report, ECF No. 841-1). Bergman’s proposed royalty would apply only to potential future harm. Bergman Deposition at 169-70. Bergman does not offer an opinion about how to determine whether any future use of the 7 CI was caused by Novelis’ disclosure in the

2016 patent application and Bergman is not qualified to do so. Bergman Deposition at 174-75. Bergman did not do any analysis to determine whether there would ever be a future use to which his proposed royalty would attach. Bergman Deposition at 177-78. Bergman recognizes that Arconic would be entitled to damages only if an unknown future user did not obtain the rights to the A951 process through Chemetall or Arconic. Bergman Report ¶¶ 142, 146; Bergman Deposition at 170-71. Bergman agreed that the parties would have to come back to court at a later date to determine whether any future use involved the 7 CI and where the user got that information. Bergman Deposition at 181. Novelis articulates numerous objections to Bergman’s opinions. The court need not address each of Novelis’ contentions. Fundamentally, Novelis points out that there are no actual

damages. Novelis argues that Bergman’s opinions do not “fit” because they would invite the factfinder to speculate about the measure of non-existent damages and would result in a windfall. Novelis also contends that Bergman did not apply a proper methodology (or any methodology at all) to calculate recoupment of R&D costs. Novelis asserts that Bergman did no independent analysis to determine R&D damages -- he simply accepted an informal estimate by former Arconic employee James Marinelli (“Marinelli”) that R&D costs were $4.9M.6 Novelis objects that Bergman’s apportionment conflicts with Marinelli’s deposition testimony and Arconic’s

6 See Bergman Deposition at 140-151. During the June 28, 2022 oral argument, Arconic asserted that an expert must provide some kind of analysis and cannot simply parrot what a fact witness said. Tr. at 91- 92. interrogatory responses that its development costs cannot be segregated into the 7 CI or estimated.

A. Harm

Bergman’s opinions omit an essential prerequisite. Bergman and Arconic forthrightly admit that there were no actual damages resulting from the alleged breach. Bergman Deposition (ECF No. 841-3) at 169 (“I haven’t seen any evidence of last profits”); Arconic Rule 30(b)(6) Deposition through designated representative David Coates (“Coates Deposition”) (ECF No. 841-5 at 339 (“At this point in time, we have not positively identified any lost profits that we can correlate to that disclosure. . . . At this point, we do not have any direct evidence that we have lost sales due to the disclosure of the seven CI”). Six years have elapsed and there is no reason to believe that damages will begin now. There can be no continuing breach by Novelis – there was a one-time disclosure in 2016. Since that time, the A951 process has been used commercially by Arconic, Novelis and MSC. The court could find no analysis or factual

background to support an argument that the initial values disclosed in the ‘440 patent application have remaining value in light of the actual commercial use. Bergman admits that he did no analysis to determine what transactions might result in future damages and he is not qualified to do so. Bergman Deposition at 174-175. Bergman does not offer an opinion about how to determine whether any future use of the 7 CI was caused by Novelis’ disclosure in the 2016 patent application. Bergman Deposition at 174-75. Bergman agreed that the parties would have to come back in the future to determine whether any future use involved the 7 CI and where the user got that information. Bergman Deposition at 181. Based upon the record developed, any future damages resulting from the 2016 disclosure would be purely speculative. Bergman concedes that any future damages could not involve Novelis or Chemetall (the exclusive distributor of A951) because Arconic gets royalties on all sales through Chemetall. Bergman Report ¶¶ 142, 146. The following essential circumstances are entirely speculative:

(1) who could possibly make a violative sale; (2) what product could constitute a violative sale; (3) whether any sale resulted from use of the 7 CI – as opposed to the other 281 claimed items of confidential information on which Arconic’s claims were stricken, or any other information; (4) whether that unknown future party independently developed or lawfully obtained the 7 CI; (5) whether Arconic could causally connect a future violative sale to Novelis’ disclosure in the ‘440 patent application in 2016; and (6) whether the 7 CI disclosed in Novelis’ 2016 patent application have any value in light of the years of real-world, actual commercial experience producing A951-treated aluminum on the production lines at Arconic, Novelis and MSC.7 Because Arconic did not demonstrate the existence of any damages caused by the alleged breach, the court need not reach any expert opinion about the amount or calculation of damages.

See Brader, 64 F.3d at 878; ATACS, 155 F.3d at 670; Kivett, 2022 WL 1185885 at *6 (refusing to award damages where there was no evidence regarding future sales); Hand, 2002 WL 227174, at *6. In short, Bergman’s opinions about how to calculate the amount of Arconic’s damages on the 7 CI8 do not “fit” for the simple reason that there are no such damages.

7 Of course, to prove a breach Arconic must also demonstrate, among other things, that the 7CI were Arconic’s information, were confidential, were disclosed in a way that Novelis had notice to keep them confidential, were not previously known to Novelis or the industry, and were disclosed by Novelis in the 2106 patent application. The Eden and Ernst expert reports address these issues. 8 To repeat, the A951 chemical formulation itself is not claimed as one of the 7 CI. B. Speculation and windfall The court need not reach Novelis’ specific methodology objections. The court will briefly explain why – even if it reached the calculation of damages -- it concludes that Bergman’s royalty and R&D recoupment opinions require undue speculation and invite the factfinder to

award Arconic a windfall. Based upon the record, the court concludes that neither theory is cognizable in this case. 1. Running royalty Bergman does not offer an opinion on royalty damages. Instead, he suggests that the court should merely determine a royalty rate, which would then be applied in a future proceeding to unidentified future sales. Bergman did not do any analysis to determine whether there would ever be a future use to which his proposed royalty would attach. Bergman Deposition at 177-78. Bergman’s royalty opinion is premature. He asks the court to impose a standard royalty rate to all future situations, without adequate information. For example, would the royalty for use of 1 of the 7 CI be the same as the royalty for use of all 7 CI? Would the royalty be affected

by how the 7 CI were obtained, how important the 7 CI were to making the sale, or the profits realized from the sale? See XpertUniverse, Inc. v. Cisco Sys., Inc., No. CV 09-157-RGA, 2013 WL 6118447, at *14 (D. Del. Nov. 20, 2013), aff'd, 597 F. App'x 630 (Fed. Cir. 2015) (denying ongoing royalties in a patent case where there were “no future sales of any infringing product upon which to base an enhanced ongoing royalty”). In effect, Arconic seeks an advisory opinion about one component of a speculative future damages calculation. The Third Circuit Court of Appeals recently criticized a similar request: “Seeking to litigate this ostensible controversy now over unfiled, potential future damages claims is the very sort of speculative, ‘hypothetical’ factual scenario that would render such a declaratory judgment a prohibited advisory opinion.” Schell v. OXY USA Inc., 814 F.3d 1107, 1115 (10th Cir. 2016) (alterations and quotation marks omitted).

Johnson v. Governor of New Jersey, No. 21-1795, 2022 WL 767035, at *2 (3d Cir. Mar. 14, 2022). In sum, Bergman’s opinion about a royalty rate to be applied to unknown future sales is simply not helpful. See Allergan Sales, LLC v. UCB, Inc., No. 215CV01001JRGRSP, 2016 WL 8222619, at *3 (E.D. Tex. Nov. 7, 2016) (“Any opinion on a future royalty that may accrue from infringement that has not yet occurred . . . is properly excludable under Rule 702(a) as being contrary to law or as not helpful to the trier of fact to determine “a fact in issue.”). 2. Recoupment of R&D costs Bergman’s R&D recoupment opinion would result in a windfall, which Arconic concedes is improper under Pennsylvania law. (ECF No. 934 at 3). The record establishes that Arconic suffered no actual damages from Novelis’ alleged breach. Arconic is being handsomely compensated for its R&D investment -- Novelis has been paying royalties to Arconic for the past six years and continues to do so. See Hanwha, 2014 WL 2452892, at *15 (plaintiff cannot recover development costs when it is still capable of manufacturing and profiting from the

product). Arconic was required by Ford to share this information with Novelis as a condition of using A951 for the F-150, a program from which Arconic has richly benefited. There are no principles of equity or unjust enrichment that would support any additional payment to Arconic in this case. Hansen, 367 F.3d at 1315 (restitution inappropriate where relief would result in an unfair windfall). In sum, the measure of restitutionary or reliance damages from Novelis’ alleged disclosure of the 7 CI is clearly $0. Arconic is clearly in a better position now than if it was returned to the position it occupied before the contracts at issue were executed. The court does not reach Novelis’ other asserted reasons to exclude these opinions. C. Conclusion about Bergman Bergman’s opinions about the 7 CI damages do not “fit.” Arconic failed to establish the existence of damages and therefore, there is no need to consider any opinions about the amount of damages. Even if the court were to reach the issue of how to calculate the amount of

damages, Bergman’s royalty and R&D recoupment theories require undue speculation and would result in a windfall. The Daubert motion to exclude Bergman’s opinions about the 7 CI damages (ECF No. 389) will be GRANTED. Arconic is entitled to recover only nominal damages if it proves a breach of contract claim against Novelis for disclosure of the 7 CI.

V. Other 7 CI experts The Eden, Novelis’ expert, and Ernst, Arconic’s expert, in their expert reports address, from a technical perspective, other elements of a breach of confidential information claim. Eden, a material sciences professor, analyzed each of the 7 CI and opined that each was: (1) not disclosed in the ‘440 patent; (2) was publicly known in the industry prior to Novelis’ disclosure;

(3) was already in Novelis’ possession; (4) was not disclosed by Arconic; or (5) Arconic failed to protect it as confidential. Arconic seeks to exclude Eden’s opinions, in part, to the extent he offers various legal opinions, i.e.,: (1) Novelis did not breach the contract; (2) the information does not constitute CI; (3) his interpretation of the PTO decisions; (4) his interpretation of the third-party NDAs; and (5) his interpretation of this court’s decisions. Ernst is an engineering professor. He offers competing technical opinions about how the 7 CI are valuable in connection with the A951 process. Novelis argues that Ernst’s opinions about meaning, interdependence and criticality are contrary to the court’s summary judgment ruling (ECF #623 at 37-40), which held that Arconic failed to properly identify any CI, except for the 7 specific items identified by the special master. In other words, Novelis contends that Arconic cannot outrun the “law of the case” limits on its recovery by saying that the court must consider the process as a whole, or that values outside the ranges set forth in the 7 CI can still be confidential.

The court made several rulings on the record during the oral argument about the other 7 CI experts. To recap, the court ruled: (1) Eden cannot opine on the legal effects of NDAs; (2) the rebuttal opinions of Stalters, Arconic’s expert, about the legal effects of NDAs, therefore, also will not be admissible; (3) Eden cannot opine on the legal effect of a PTO ruling or this court’s rulings; (4) Eden and Ernst will be bound by the same standard – each can explain technically what was happening and what a POSITA would understand (for example, whether ion exchange is an optimization process), but they cannot opine about whether or not there was a breach of contract; and (5) Eden and Ernst can testify about whether the 7 CI are technically important, but not whether they are “valuable.” Counsel were directed to meet and confer about appropriate verbiage to be used during trial. See June 28 Transcript; Minute Entry of June 29,

2022. The parties do not challenge the technical opinions offered by Eden and Ernst. The disputes involve the extent to which the experts may offer legal opinions or opinions going to the ultimate issues in dispute. Pursuant to Federal Rule of Evidence 704, an opinion on an ultimate issue is not automatically objectionable and the court has discretion to admit such testimony. Experts, however, cannot offer legal opinions or usurp the court’s role. In utilizing that discretion, however, the District Court must ensure that an expert does not testify as to the governing law of the case. Although Federal Rule of Evidence 704 permits an expert witness to give expert testimony that “embraces an ultimate issue to be decided by the trier of fact,” an expert witness is prohibited from rendering a legal opinion. United States v. Leo, 941 F.2d 181, 195–96 (3d Cir. 1991). Such testimony is prohibited because it would usurp the District Court's pivotal role in explaining the law to the jury. First National State Bank v. Reliance Elec. Co., 668 F.2d 725, 731 (3d Cir. 1981) (per curiam).

Berckeley Inv. Grp., Ltd. v. Colkitt, 455 F.3d 195, 217 (3d Cir. 2006). As the court stated on the record, Eden’s interpretations of the PTO decisions, NDAs and this court’s decisions would be improper and will not be admissible. Eden will not be permitted to opine about the ultimate issues whether or not Novelis breached the contract or whether the 7 CI constitute “confidential information.” Eden will be permitted to offer factual, technical opinions on those subjects. Indeed, Arconic does not object to those technical opinions. An equal rule will apply with respect to Ernst’s opinions – he can offer technical opinions about the importance of the 7 CI, whether the 7 CI were confidential, whether they were disclosed to Novelis and whether Novelis disclosed them in the ‘440 patent. Novelis does not object to those opinions. Neither expert, however, will be permitted to offer opinions on the ultimate issues whether or not Novelis breached the contract or whether the 7 CI constitute “confidential information.” Counsel must meet and confer on the appropriate verbiage to be used at trial to resolve these disputes. Ernst will not be permitted to offer opinions that are contrary to the court’s summary judgment ruling (ECF #623 at 37-40). It is law of the case that Arconic cannot recover for any CI, except for the 7 specific items identified by the special master. Arconic will not be permitted to invite the jury to consider: (1) the process as a whole; (2) the “fundamental meaning” about

the underlying physics; or (3) that numbers outside the ranges set forth in the 7 CI can still be confidential. Ernst can opine only on the ranges set forth in the 7 CI9 and cannot invite the jury to decide the claim on any other basis.

9 Ernst can opine, from a technical perspective, why the particular ranges disclosed in the 7 CI are important, e.g., that a certain concentration level of a particular parameter is necessary to achieve complete deoxidation. Expert testimony with respect to whether or not the 7 CI must be evaluated in the context of the A951 process will be permitted. Both Eden and Ernst can opine, from a technical perspective, about how knowledge about the 7 CI in the context of a different pretreatment process impacts confidentiality (or lack thereof) in the context of the A951 process.

If Arconic’s 7 CI claims are subject to a bench trial, the court can give the remaining expert opinions relating to the 7 CI the appropriate weight. The court may provide more detailed rulings, upon motion of the parties, if count 4 survives summary judgment and will be tried by jury.

VI. Conclusion In summary, for the reasons set forth above, the Daubert motion to exclude Bergman’s opinions with respect to the 7 CI (ECF No. 839) will be GRANTED; the Daubert motion with respect to Stalters (ECF No. 852) will be GRANTED, as stated on the record; and the Daubert motions with respect to Eden and Ernst (ECF Nos. 849, 861, 864) will be GRANTED IN PART

and DENIED IN PART without prejudice as stated above and on the record. An appropriate order follows. Dated: November 18, 2022 By the Court: /s/ Joy Flowers Conti Joy Flowers Conti Senior United States District Judge