Copperweld Steel Company, in No. 77-1493 v. Demag-Mannesmann-Bohler, Demag Stranggiess-Technik Gmbh, Demag Aktiengesellschaft, in No. 77-1494

OPINION OP THE COURT ROSENN, Circuit Judge. Plaintiff, Copperweld Steel Company (“Copperweld”) appeals from a final judgment of the district court of the Western District of Pennsylvania in favor of defendants, Demag-Mannesmann-Bohler, Demag Stranggiess-Technik GmbH, and Demag Aktiengesellschaft (collectively “Demag”) and requests that we order a new trial because of prejudicial errors committed at trial. Defendants cross-appeal from the district court’s exercise of jurisdiction over the action and refusal to dismiss the case pursuant to a forum selection clause providing for the resolution in Germany of any disputes arising from a contract between Copperweld and Demag. They ask for a new trial on the question of damages arising from the failure of the court to enforce this clause of the contract. This is a diversity suit based on four causes of action: (1) breach of contract, (2) negligent design and manufacture, (3) negligent misrepresentation, and (4) fraudulent misrepresentation. At trial Judge Barron P. McCune denied Demag’s motion to dismiss the action, directed a verdict against Copperweld on the fraud claim, and entered final judgment on the jury’s verdict against Copperweld on the other claims. We affirm. I. In the early 1960’s, Copperweld was producing special high quality carbon and alloy structural steels at its Warren, Ohio plant. In order to remain competitive in this specialized business, the company recognized a need to improve its manufacturing facilities at Warren by either increasing productive capacity or reducing costs. To achieve this end, it entered into negotiation with several companies, including Demag, for the purpose of purchasing new and more cost efficient facilities for the production of high quality steels. Copperweld’s Warren plant fashioned its steel by the ingot casting method by which molten steel was poured into a number of ingot molds, allowed to cool slowly, and eventually to solidify into a large chunk or ingot of steel. The ingot would then be cut from the mold, resulting in some loss of structurally sound steel and leaving an unusable chunk of metal. To obtain a salea-ble product, further processing, therefore, was necessary. At this point, the steel was reheated and passed through rollers — like those of an old-fashioned washing machine — to reduce the slab of steel into a semi-finished product called a billet (steel bar). This additional processing produced further waste through the reduction in size and cutting of the ingot. Furthermore, casting by this method also proved to be inefficient, for no steel could be cast until the prior ingot was entirely finished and the time required to fill any order for usable steel, consequently, was quite lengthy. To eliminate these costly steps, decrease metal loss, and decrease production time, Copperweld sought to obtain a more efficient machine — a continuous casting facility. Continuous casting increases the yield of usable steel by allowing the direct fabrication of billets from molten metal poured into a casting mold, thereby eliminating the intermediate steps of ingot casting. This is accomplished in the following manner: molten steel is poured into a small, water-cooled mold, the bottom of which is covered by a plug which may be pulled from the mold. When the liquid metal is poured into the machine, its outer shell begins to solidify as a result of contact with the water-cooled walls of the caster. Similarly, the molten steel begins to freeze to the plug at the bottom of the mold. As more metal is poured into the container, the plug at the bottom is slowly withdrawn from the mold and sprayed with water to solidify the core *956 while another portion of liquified metal is poured into the now vacated mold. This process continues until all of the available molten metal is poured. After the molds solidify, they are cut into usable lengths and placed on a hotbed to cool. Because of the opportunity to save both time and expense, Copperweld solicited various bids for a continuous casting system. It received its most serious proposals from Concast, Inc., and Demag, two German designers and manufacturers of steel making facilities. Apparently, both of the bidders offered continuous casting machines, but of radically different designs. Demag offered a straight mold system which would cast the steel strands vertically until they would reach the desired length; they would then be bent into a horizontal plane, finally cut, and then removed from the system. Con-cast, however, was ready to offer a low head curved mold system which would cast the steel horizontally because the molten liquid would be poured into a curved mold and automatically forced to leave the caster in a horizontal position. The straight mold caster suffers distinct disadvantages. It requires a highly expensive and tall building in order to cast the steel at an appropriate height to make the vertical mold. The procedure is not only costly — requiring most steel companies to construct new tall buildings — but dangerous because raising the molten steel to the desired height could induce spillage. The curved mold design alleviates such problems by initiating the steel into the mold in a curved position, thus obviating the need for buildings with high ceilings. Copperweld, for these obvious reasons, preferred the curved mold low head facility over the straight mold. It informed Demag of this preference. Demag’s internal mem-oranda confirm that its American agent relayed Copperweld’s inclination toward the curved mold to Demag and also informed Demag that if it desired to acquire the lucrative Copperweld contract it would either have to make a curved mold offer or demonstrate to Copperweld the superiority of the straight mold. Demag chose to continue its recommendation of the straight mold design and specifically informed Cop-perweld that the reliability of the curved mold system was questionable, but that its capabilities would be established some time in the following year, after testing of an experimental facility by Demag’s research affiliate. Shortly thereafter, Demag submitted a formal proposal to Copperweld for a vertical casting machine, stressing that the data was insufficient to enable it to recommend a curved mold. However, approximately two weeks later, Demag abruptly changed its proposal and instead offered a curved mold design. Copperweld accepted this proposal which became the basis for the contract sued upon in this case. No documented explanation exists for this sudden shift of position by Demag and the parties suggest two radically different rationales for the action. Copperweld contends that Demag was forced to alter its proposal to forestall the award of the contract to Concast, asserting that Demag went ahead with its curved mold proposal even though it had no reason for confidence in the reliability of the design. Demag maintains, however, that its proposal came only after extreme pressure by Copperweld to produce a curved mold system. Its position is that it fully informed Copperweld of the risks of the curved mold system, but that it moved forward with the proposal because of the acceptance of the risks by Copperweld’s Executive Vice President, C. W. Holmquist, an acknowledged expert in the steel casting business. 1 *957 No one integrated document makes up the contract between Demag and Copper-weld. Although the parties issued various proposals, purchase orders, and correspondence to each other, consolidation of the documents reveals certain elements of the contract with clarity: (1) that the machine was to be designed by Demag, (2) that the machine actually was to be constructed in the United States by Birdsboro Corporation, although it originally was to be manufactured in Germany by Demag, (3) that the machine was warranted as capable of producing 95-97 percent sound strands of steel at a rate of twelve heats per day, 2 (4) that a German crew would operate the caster initially, eventually training Copperweld employees to take control of the operation, and (5) that during this break-in period, various technical problems would be expected which Demag would be required to cure. Copperweld brought this suit asserting that the machine did not meet, nor was it capable of meeting, the warranted production guidelines regarding yield and number of heats. The record reveals that the machine was made operational on October 26, 1965, and that it ran at various times up to September of 1969 with little success. Cop-perweld instituted two campaigns to test the capability of the caster, the first from the day the machine because operational until July of 1967 and the second from the winter of 1969 until September of that year. It is undisputed that the caster did not operate satisfactorily to Copperweld. It is disputed, however, whether the machine was capable of meeting the specified production requirements and if it was not, whether this was the fault of Copperweld or of Demag. The district court found no evidence sufficient to go to the jury on the question of whether Demag had fraudulently misrepresented the capabilities of the machine and directed a verdict for Demag on this count of the Copperweld complaint. In response to specific interrogatories, the jury expressly found no negligence in design or manufacture by Demag, no negligent misrepresentations by it, and no breach of contract. Copperweld asserts in this appeal that the district court erred in its charge to the jury regarding breach of contract, negligent design, and negligent misrepresentation. It also asserts that the directed verdict on fraudulent misrepresentation was erroneous and that the court committed various errors in the trial of the case. We consider these claims seriatim. II. Copperweld asserts that it is entitled to a new trial on its breach of contract claim against Demag because the district court’s disjointed and confusing charge prevented fair litigation of the claim. Although the charge given was not a model of clarity, under the plaintiff’s theory of breach of contract we cannot conclude that the instructions given were erroneous. Copperweld’s basic theory at trial was that Demag breached the contract by its *958 failure to design a continuous caster capable of producing at 95-97 percent yield of sound strands at 12 heats per day. 3 Cop-perweld contends that the district court erroneously charged the jury under this theory (1) because it failed to give an adequate legal definition of breach of contract, (2) because it confused the standards of breach of contract with those for negligent design and manufacture, (3) because it incorrectly stated the definition of yield of sound strands, and (4) because it stated the conditions of the contract in a confusing and misleading manner. Demag asserts that the charge in its entirety was fair to Copperweld and that Cop-perweld’s contentions on appeal, other than those concerning the definition of yield and the conditions of the contract, were not properly preserved. 4 From this they reason that the district court’s charge must be sustained. We agree. (A) Jury Instructions On Breach of Design Services Contract On its breach of design services contract theory, Copperweld requested that the court charge the jury that under the contract Demag was legally obligated to design and provide a low head curved mold continuous caster which . . . was capable of producing ... a yield of usable billets as compared by weight to the steel in the ladle brought to the continuous caster in the range of 95-97% . at a production rate of 12 heats per day. (Emphasis supplied.) Copperweld further asked the court to instruct the jury that if they found that the caster was not capable consistently either of producing a yield in the 95-97 percent range or of operating at 12 heats per day “then that is a finding of a breach by Demag of the contract between Copperweld and Demag.” *959 It is apparent from these requested points for the charge that Copperweld’s breach of contract claim was based totally on the capacity of the continuous caster to produce according to the warranted figures for yield and number of heats. A review of the court’s charge on this claim of breach of contract reveals that although it is not identical to that requested by Copperweld, it essentially conveyed the elements of Cop-perweld’s theory on breach. Consequently, in the admitted absence of any specific objection to this part of the charge by Copper-weld after the charge was read to the jury, we find no reversible error. After explaining to the jury all of the contentions of the parties, outlining the proof at trial, and defining yield 5 and the other critical terms in the contract, the court charged the jury as follows: The plaintiff . . . has brought suit alleging that Demag breached a contract it had entered into with Copper-weld. As I understand it, the allegation is that the continuous caster would not perform in accordance with the terms of the contract. * # Of sis * $ . The real issue here is . . . did defendant breach the contract? Did defendant design a curved mold machine which was incapable of meeting the terms of the agreement even if the design was all right? Even if the design was all right, was the machine manufactured so poorly that it could not meet the terms of the agreement? Let me go back over that because I think I stated that poorly. The issue . .is: Did defendant breach the contract? Did the defendant design a curved mold machine which was incapable of meeting the terms of the agreement, or, even if the design was all right, was the machine manufactured so poorly that it could not meet the terms of the agreement? In essence, the court’s charge on breach of contract and that requested by Copper-weld rest essentially on the capability of the machine to fulfill the warranted contract terms. That the court did not tell the jury specifically that lack of capacity of the caster to meet the yield and number of heats requirements, as warranted, constituted a breach of contract, is not error because the charge itself equated the incapability of the machine of “meeting the terms of the agreement” with breach of contract. Therefore, under Copperweld’s theory of contract liability, the charge is not erroneous. 6 (B) Confusion Between Breach of Contract Standards And Negligence Theories Copperweld also asserts that the charge was incorrect because it confused plaintiff’s theories of liability — breach of contract, negligent design and manufacture, and negligent misrepresentation — so that the jury could not fairly decide the contract question. This assertion is based on the language in the charge, cited above, which instructed the jury that breach should be found if either the design of the caster rendered it incapable of performing as warranted or if the manufacture of the caster rendered it incapable of performing as war *960 ranted. This instruction, Copperweld asserts, required the jury to find fault on the part of Demag as a condition precedent to liability. Copperweld insists that this is tantamount to equating breach of contract with negligence, when coupled with the following instruction also given by the court: It seems to me that your [jury’s] answers to the questions [the interrogatories on Demag’s liability under Copper-weld’s different theories] should be consistent, although they need not be. . . . For example, if you find that the contract was not breached, then it seems to me Demag was not negligent in designing or manufacturing the continuous caster because you would have found it capable of doing what Demag said it would do. Further, it would have been no negligent misrepresentation because the caster would have been found capable of doing what Demag said it would do. On the other hand, if you find that the contract was breached, then the caster you would have found was not capable of doing what Demag said it would do, and in that case Demag may well have been guilty of negligence in design or manufacture of the caster . The district court intertwined some aspects of negligence in its charge on breach of contract, but the court specifically informed the jury that it was free to find inconsistency between the theories. Moreover, we believe that Judge McCune, on the whole, fairly described each of Copper-weld’s independent theories of liability, thereby minimizing the potential confusion to be caused by the blurring of distinctions between the individual claims and that he correctly perceived Copperweld’s case rested on one basic factual assertion — that the caster, as designed, was incapable of performing as warranted. That this same fact is also the predicate for each of Copper-weld’s independent theories of liability is itself the probable basis for the confusion of which Copperweld now complains. In a conference prior to the charge, the district court advised counsel of its intended instructions on breach of contract, presaging those eventually given: You [Copperweld] say they [Demag] breached the contract because the machine would [not] do what they said it would do with respect to yield and frequency of operation, and the reason they [Demag] breached the contract was because the state of the knowledge was such that they couldn’t design a machine that would do what they said it would do and therefore they didn’t as a matter of fact accomplish it, and, secondly, the machine wasn’t properly made. After the court read this proposed charge, Copperweld expressed concern that the instructions on breach might confuse the jury as to the difference between breach of contract and negligence. The court noted these general objections and proceeded to give the instructions quoted in part 11(A), supra. However, at the close of its charge, the court asked for specific objections from counsel. Copperweld failed to object to the general charge on breach, objecting only to certain portions of the instruction, see part 11(C), infra. Instead it chose to rest on its automatic exception given for the denial of its requested charge. We conclude that Copperweld’s requested instruction was fairly covered by the district court and that Copperweld’s objections on appeal were not otherwise properly preserved. Fed.R.Civ.P. 51. 7 We therefore find no error because of any confusion in the charge. (C) Definitional Instructions On “Sound Strands” Copperweld, at the close of the court’s charge, specifically noted dissatis *961 faction with the court’s instructions with regard to yield of sound strands and with regard to certain performance conditions on Copperweld contained in the contract. An assessment of the court’s charge and Cop-perweld’s objections thereto convinces us that the instructions were not erroneous. The keystone of Copperweld’s breach of contract case was that the caster was warranted as capable of producing 95-97 percent yield of sound strands. The precise definition of this warranty was one of the most hotly contested issues at trial. Demag contended that the number of sound strands was to be determined by comparing the weight of sound billets on the cooling bed after casting with the total weight of molten metal poured into the caster. Cop-perweld contended that the term was to be defined by reference to the weight of usable metal remaining after cooling and inspection compared to the total weight of the molten steel. Both Demag and Copper-weld agreed, however, that the parties to the contract contemplated that the warranty referred to billets that would be “usable” to Copperweld. Although the testimony at trial varied as to the correct definition of sound strands, the court charged the jury favorably to Demag: It is clear what this phrase [yield of sound strands in the region of 95 to 97%] meant and I will give you a definition of that phrase because I think you must have that definition as well in order to settle the issues in this suit. The definition of yield is this. Yield means the weight of crop billets taken from the machine, that is billets on the cooling bed [before inspection], compared to the weight of the liquid poured into the casting machine. Copperweld contends that the court’s apparent acceptance of Demag’s definition of yield is reversible error because, at a minimum, the evidence at trial presented a jury question as to the exact contractual meaning of the term “sound strands.” Had the district court said nothing more about the definition, we would agree. 8 The court, however, went on to explain that the billets — that is the cropped billets which come off the machine were intended by both parties to be useful. It was intended by both parties that these billets were to be used to sell to a customer . Whether the machine could meet the yield figures or meet them at a rate of twelve heats per day and whether the billets could be sufficiently useful are points in contention in this lawsuit. (Emphasis supplied.) Read in context, this addition to the definition of “sound strands” given by the district court fully and adequately presented to the jury the critical question concerning Demag’s liability for breach of contract — whether or not the yield was useful to Copperweld. Given such an instruction, the jury was told the crux of Copperweld’s claim. The jury could fairly decide whether Demag had breached its warranty. Moreover, this part of the court’s charge did no more than repeat Copperweld’s requested point that the machine be “capable of producing . . . yield of usable billets.” In view of Copperweld’s failure to ask for a specific point on the meaning of “usable,” the refusal of the court to instruct the jury that they should construe yield of usable billets as those remaining after inspection is harmless error at best. We therefore perceive no reversible error in the court’s instructions on “sound strands.” 9 *962 III. Copperweld contends that the district court erred in directing a verdict in favor of Demag on the fraudulent misrepresentation claim. 10 Copperweld asserts that there was sufficient evidence of fraud to have gone to the jury and that a new tjrial on the issue must be given. We agree with Copperweld that the evidence presented a factual question as to fraud. Viewing the evidence favorably to Copperweld, as we must on appeal from a directed verdict, see Hunziker v. Scheidemantle, 543 F.2d 489, 494 n. 8 (3d Cir. 1976); Dougherty v. Hooker Chemical Corp., 540 F.2d 174, 178 (3d Cir. 1976), we believe that sufficient evidence exists to support the claim. The jury could have found the following: (1) During 1962, Demag and Copperweld began negotiating over construction of a continuous casting machine. Copperweld sent Demag samples of the type and quality of metal that the machine would have to be capable of casting. (2) By May 17, 1963, Demag had committed itself to offering Copperweld a straight mold caster. Yet, at this time it asked its research affiliate to undertake a study of whether a curved mold facility could satisfactorily produce steel to Copperweld’s specifications. Copperweld was informed that tests were being undertaken by Demag on the feasibility of a curved mold, but that the tests had not yet provided a basis to recommend such a system. (3) On May 21, 1963, Demag submitted a proposal for a straight mold caster. Demag also learned that its competitor, Concast, Inc., had submitted a formal proposal for a curved mold caster. (4) On June 21, 1963, Demag’s United States agent informed Demag that Copper-weld preferred a curved mold design similar to that proposed by Concast. Demag was told that to obtain the contract it either had to recommend a similar curved mold machine or had to establish the superiority of the straight mold caster. (5) On June 25, 1963, Demag sent a telex to its research affiliate requesting that it take a firm position as to the capability of the curved mold caster to produce steel to Copperweld’s specification. Demag, however, was informed that such a system could reliably produce only certain steels. (6) On June 26, 1963, Demag’s United States affiliate quoted Copperweld a price for a straight mold caster and repeated Demag’s reservations about the reliability of the curved mold design. One day later, these warnings were again repeated. (7) On July 11, 1963, only two weeks after proposing a straight mold caster and issuing warnings to Copperweld about the lack of reliability of the curved mold design, Demag made a formal proposal for the curved mold machine which has become the focus of this lawsuit. *963 (8) From these events, the jury could have reasoned that Demag, fearing loss of the lucrative Copperweld contract, made a proposal for a curved mold caster without a reasonable basis to expect that its representations were in fact true. Such recklessness could have provided the predicate for a finding of fraudulent misrepresentation. Although the evidence and the inferences to be drawn therefrom demonstrate a factual question which normally should have been sent to the jury, we conclude that the failure to do so was harmless here because the jury expressly found that Demag did not even negligently misrepresent the capacity of the caster. 11 Furthermore, the jury’s express finding that Demag did not breach its contract, see part II, supra reinforces our conclusion that even if Demag may have made misrepresentations, 12 they were not material and did not involve the critical qualities of the caster. Therefore, the court’s directed verdict on the fraudulent misrepresentation claim was not reversible error. IV. Copperweld complains that the district court committed various evidentiary errors in the trial of this case. Only one of these claims warrants discussion — whether the court erroneously admitted a hearsay memorandum prepared by a Copperweld attorney — as we find the other claims without merit. 13 Copperweld maintains that the admission of the memorandum, consisting of a Copperweld attorney’s recollection of an interview with C. W. Holmquist, the former Executive Vice President of Copperweld’s steel division, violated the hearsay rule. Controversy over this evidence first arose during discovery, when the trial court ordered Copperweld to turn over to Demag any items generated during the course of the litigation involving reports and statements of C. W. Holmquist. The plaintiff opposed this order on a work product theory, but the court rejected the contention and ordered production of the documents. 14 At trial, the district court, on Demag’s motion and over Copperweld’s objection, admitted the Holmquist memorandum under rule 804(b)(5) of the Federal Rules of Evidence as an exception to the hearsay prohibition. Rule 804(b)(5) provides that an out of court statement, with equivalent guarantees of trustworthiness as other hearsay exceptions in the federal rules, is admissible if offered as evidence of a material fact, so *964 long as the statement is more probative on the point for which it is offered than any other evidence readily procurable and its admission would serve the general purposes of the rules of evidence. 15 The trial court found the memorandum admissible under the rule: (1) because it was material, “stating the attitude of the person who was most clearly involved in the dispute”; (2) because it was trustworthy, there was no reason presented to doubt its truth and the statement was authenticated as being that of- Holmquist; 16 (3) because it was more probative than any other evidence admitted 17 as it was direct evidence of the thoughts of Holmquist, Copperweld’s officer in charge during the negotiations and early trials with the caster; and (4) because it was reviewed by Holmquist and apparently adopted as his statement of the case, it would serve the interests of justice as well as the rules to admit the memorandum. Copperweld attacks these findings as insufficient to support the admission of the memorandum. In reviewing the trial court’s admission of this evidence we are called upon to assess both a conclusion of law, that the memorandum was admissible, and findings of fact, trustworthiness, materiality, and the like. In review of the factual underpinning for the admission of the evidence, we must decide only if the findings of fact are clearly erroneous. See United States v. Bagley, 537 F.2d 162, 166 (5th Cir. 1976) (court of appeals’ assessment of district court’s finding of trustworthiness in the declaration against interest exception to the hearsay rule, reviewed by the clearly erroneous standard). Under that standard there is ample support for the findings. See United States v. Ward, 552 F.2d 1080, 1082 (5th Cir. 1977) (hearsay oral testimony corroborating that stolen vehicles were in interstate transit admitted under 804(b)(5) as trustworthy even though declarant had strong motivation to falsify the information). We therefore hold that admission of the document was not error. V. Demag has raised a single but interesting question in its cross-appeal— whether the district court erred in accepting jurisdiction in this case. Demag claims that the district court did err and that this error entitles it to a trial on the question of damages suffered because of the district court’s failure to enforce the contract. We find no error in the retention of jurisdiction and therefore affirm the district court’s disposition of Demag’s claim. During negotiations Demag sent Copper-weld a standard form with its conditions for export contracts. Among these conditions was one requiring that “[a]ny disputes arising out of the terms of the contract” would have to be brought before a German court unless Demag chose to bring the action in the United States. This condition expressly *965 applied to any export contract negotiated by Demag. Prior to the final agreement on the sale of the caster, however, Demag and Copperweld concluded that the machine, originally to be built in and exported from Germany, would be manufactured in the United States. From this Copperweld reasons that the agreement ceased to be an export contract, that Demag’s forum selection clause did not become part of the new contract, and that therefore the district court correctly retained jurisdiction over the case. The district court indicated some agreement with Copperweld’s assertion, concluding that had the parties not amended their contract to provide for construction of the caster in the United States, it would have been inclined to enforce the forum clause. We need not decide whether the clause in fact became part of the contract, however, for there is an alternative reason that leads us to affirm the district court’s assumption of jurisdiction over the case. The district court also reasoned and concluded that enforcement of the forum selection clause would be unreasonable. Copperweld Steel Co. v. Demag-Mannesmann-Boehler, 354 F.Supp. 571, 572-73 (W.D.Pa.1973); 347 F.Supp. 53, 54-55 (W.D.Pa.1972); 54 F.R.D. 539, 542 (W.D.Pa.1972). We see no error in that conclusion. During pre-trial motions, the district court held that it had jurisdiction over Cop-perweld’s complaint and that the forum selection clause did not deprive it of this jurisdiction. Citing Central Contracting Co. v. Maryland Casualty Co., 367 F.2d 341 (3d Cir. 1966), and Central Contracting Co. v. C. E. Youngdahl & Co., Inc., 418 Pa. 122, 209 A.2d 810 (1965), Judge McCune concluded that enforcement of the forum selection clause would be “unreasonable under the facts of this case” and refused to enforce the clause. Copperweld Steel Co. v. De-mag-Mannesmann-Boehler, supra, 54 F.R.D. at 542. 18 Subsequent to this pre-trial decision, in The Bremen v. Zapata Off-Shore Co., 407 U.S. 1, 92 S.Ct. 1907, 32 L.Ed.2d 513 (1972), an admiralty case, the Supreme Court held that a forum selection clause in a contract between the parties is “prima facie valid and should be enforced unless enforcement is shown by the resisting party to be ‘unreasonable’ under the circumstances.” 407 U.S. at 10, 92 S.Ct. at 1913. The court further stated that such a clause should not be set aside “absent a strong showing” such as that enforcement “would be unreasonable and unjust, or the clause was invalid for such reasons as fraud or overreaching.” Id. at 15, 92 S.Ct. at 1916. Demag contends that the district court’s pre-trial decision that enforcement would be unreasonable must be reversed because it was based, at least in part, upon standards enunciated in the Central Contracting Company cases, cases decided prior to The Bremen. Demag also made this argument to Judge McCune, who in two separate opinions reiterated his initial reaction and held that the enforcement of the clause would be unreasonable. See Copperweld *966 Steel Co. v. Demag-Mannesmann-Boehler, supra, 354 F.Supp. at 572 (“After further consideration we believe the alleged forum selection clause to be manifestly unreasonable . . . .”); 347 F.Supp. at 54 (“We remain convinced that in this particular case plaintiffs could, not adequately try their case if forced to proceed in Germany . .”). We likewise are unconvinced that the standards applied by Judge McCune were erroneous in light of The Bremen and we cannot say he abused his discretion under the particular facts of this case. The district court’s reliance on the Central Contracting Company cases is not reversible error. Both cases held that a court should generally enforce a forum selection clause unless the enforcement would be unreasonable at the time of the litigation, Maryland Casualty Co., supra, 367 F.2d at 344; C. E. Youngdahl & Co., Inc., supra, 418 Pa. at 133-34, 209 A.2d at 816, and that mere inconvenience would not show unrea-sonability. Id. The courts instead suggested that the test is whether enforcement “will put one of the parties to an unreasonable disadvantage and thereby subvert the interests of justice.” Maryland Casualty Co., supra, 367 F.2d at 345. 19 Although The Bremen has language somewhat different than that relied upon by the district court, compare The Bremen, supra, 407 U.S. at 18, 92 S.Ct. 1907 (deny enforcement of the clause if resisting party will be effectively denied his day in court) with Maryland Casualty Co., supra, 367 F.2d at 345 (deny enforcement if interest of justice is subverted), the district court captured the essence of the Supreme Court’s opinion — the resisting party must prove unreasonability — and in fact held that Copperweld might well have been prevented from receiving “a fair and complete hearing” had the forum clause been enforced. Copperweld Steel Co. v. Demag-Mannesmann-Boehler, supra, 347 F.Supp. at 54. We therefore find no material difference between the standards applied by the district court and those requested by Demag and affirm the district court’s retention of jurisdiction. VI. For the reasons stated in this opinion, the judgment of the district court entered on the jury’s verdict and the judgment entered on the district court’s directed verdict against Copperweld will be affirmed. De-mag’s cross-appeal will be denied. Footnotes [1] Both Demag and Copperweld concede that C. W. Holmquist was the key figure in the negotiations between them. He was a pioneer in the field of continuous casting and his testimony would have been critical at the trial. He did not testify, however, because sometime shortly after the filing of this lawsuit he and his wife were murdered during an apparent robbery attempt. Because Holmquist did not testify at trial and his role in the contract negotiations was so crucial, the parties introduced the fact of his death. Copperweld asserts that Demag went far beyond that simple fact and purposely in *957 jected the circumstances of Holmquist’s death to create intrigue and prejudice against Cop-perweld. We find this assertion without merit. Although it was not necessary for Demag to give the specifics of the deaths of Holmquist and his wife, their introduction did not substantially prejudice Copperweld. To fill the void left by Holmquist’s death, Demag introduced a memorandum prepared by a Copperweld attorney which stated Holm-quist’s views on the Demag/Copperweld contract. We discuss the admissibility of that memorandum in part IV, infra. [2] The definition of sound strands and heats per day was a critical issue at trial with Copper-weld contending that Demag had breached the contract because the machine was incapable of operating at that rate. Both parties agreed that the number of heats was to be measured by the number of times that the molten liquid steel could be poured into the caster per day. Copperweld contended that measurement of sound strands required comparison between the weight of usable billets after inspection with the weight of the molten liquid poured, but Demag contended that the appropriate comparison was between weight of the billets prior to inspection and the weight of molten metal. The trial court apparently charged the jury along Demag’s line, but instructed the jury, favorably to Copperweld, that they would have to find the billets usable. See part 11(C), infra. [3] See n.2, supra. This theory developed during the course of the trial. In its complaint, Cop-perweld alleged that it was damaged as a “result of the failure of the plant ... to achieve the production and performance standards contracted for” and that it was damaged by Demag’s design of the plant which was not in accord with the contract and “was not capable of performance in accordance with . production and performance standards.” The complaint evidenced a two-pronged theory of contract liability asserted by Copper-weld: first, that the machine did not perform as warranted, and second, that its design rendered it to be incapable of performing as promised. The first theory focuses primarily on the qualities of a contract for the sale of a good; the second theory is tied to a view that the contract was for plant design services. The difference in the two theories is critical for each involves a different type of contract; contracts under the first theory are subject to a shorter statute of limitations than those under the second. Throughout the trial Demag stated that the breach of contract actions should have been dismissed because they were not brought within the statute of limitations. They asserted that this contract was one for the sale of goods and that therefore the contract was governed by the 4 year statute of limitations fixed by the Uniform Commercial Code. See 12A Pa.Stat. Ann. §§ 2-102, 2-725 (1970). Copperweld, however, claimed that the contract was primarily one for design services, outside the scope of the Uniform Commercial Code. As such, they stated that the contract was subject only to the Pennsylvania 6 year statute of limitations for written contracts. See 12 Pa.Stat.Ann. § 31 (1953). We need not resolve this conflict because of our disposition of the breach of contract claim. We note, nonetheless, that Copperweld’s approach to this case, stressing the design feature of the contract was enough to persuade the district court that the contract was not a sale of a good subject to the Uniform Commercial Code. Copperweld cannot now be heard to complain that the court, following Copper-weld’s trial strategy that this was a design service contract, unfairly restricted its charge to that theory. [4] Prior to charging the jury, the district court preliminarily informed counsel of the content of the charge. Basically, the court outlined the instructions eventually given. Counsel for Copperweld expressed some disagreement with the proposed charge to the court. Later, after the actual charge was given, the court asked for specific objections, stating that it would grant an automatic exception to any requested points that it denied. Copperweld then objected only to the court’s definition of yield of sound strands and discussion on performance conditions. Apparently, none of its other general objections were voiced. As to those points, Copperweld must rely on the automatic exception for its denied requests. See Fed.R. Civ.P. 51 and text accompanying n. 8, infra. [5] Copperweld contends that the charge on this definition was also erroneous. See part 11(C), infra. [6] We note that the charge as given may have presented substantial problems had Copper-weld continued to stress its argument that De-mag breached the contract by providing a caster which did not perform as warranted. The charge given was limited solely to Demag’s failure to design a caster capable of performing as warranted. Under that charge, the jury would not have been permitted to find a breach of contract even if it found that the machine did not perform as warranted, unless it also found that the machine was incapable of so performing. Such a limited instruction might have been erroneous had Copperweld specifically asked the court to charge on the failure of the machine itself to perform. However, Cop-perweld — perhaps fearful of the statute of limitations — requested an interrogatory for the jury only as to the capability of the caster. We therefore see no error in the limited instructions given. [7] Rule 51 provides in pertinent part that: [a]t the close of evidence . . any party may file written requests that the court instruct the jury on the law as set forth in the requests. ... No party may assign as error the giving or the failure to give an instruction unless he objects thereto before the jury retires to consider its verdict, stating distinctly the matter to which he objects and the grounds of his objection. [8] Moreover, even if we agreed that the definition of “sound strands” should have been submitted to the jury, we would nonetheless affirm because Copperweld never demanded that the issue go to the jury. Under Fed.R.Civ.P. 49, Copperweld is now precluded from contesting Judge McCune’s definition of “yield,” which is not clearly erroneous. [9] Copperweld also contends that the court’s charge on the conditions precedent to performance of Demag’s warranties was misleading and erroneous. The court instructed the jury that [t]he performance guarantees in the contract between Copperweld and Demag were conditioned upon the steels being perfectly melted and fed into the casting machine with *962 a good degree of fluidity and at the specified temperatures and upon the caster being correctly operated and serviced. If you find that these conditions were not satisfied, you must find that Demag did not breach the performance guarantees. Counsel for Copperweld objected to the use of the term “perfect” as any part of the condition, and asked the court to explain that the condition be considered within the context of a steel plant and that “perfect" in that context could not mean perfection in any absolute sense. We believe that though such a caveat would have been useful, in view of the lengthy trial, the jury had a comprehensive picture of steel plant operations and must have known that the term was not used in an absolute sense. No one argued at trial that ultimate perfection was necessary, only that Copperweld was required to fulfill its promise to operate the caster effectively and accurately. Therefore, any further explanatory instruction by the court was unnecessary. [10] Copperweld also asserts that the district court’s charge on negligent design and manufacture and on negligent misrepresentation was erroneous. Copperweld contends that the charge was marked by vagueness, and that the failure of the court to adequately explain the meaning of “state of the art,” and its failure to specifically relate Copperweld's allegations to the legal criteria for negligence and tie these claims to the charge, prevented the jury from fairly deciding the claims. We have reviewed the charge in its entirety and find it an adequate, though general, statement of the law of negligence. The jury’s verdict on both the negligent design and manufacture and on the negligent misrepresentation claim need not be disturbed. [11] See n. 9, supra. Although there is evidence in the record which could have supported a jury finding of recklessness, a jury finding that Demag had not negligently misrepresented implies a fortiori that the misrepresentations, if any, were not reckless. Moreover, there is no evidence in the record to support a jury finding of intentional misrepresentation. [12] The jury, in answering interrogatories posed to it on the question of negligent misrepresentation, stated that Demag had made misrepresentations to Copperweld. However, in view of the court’s charge on breach of contract and Copperweld’s theory on breach and misrepresentation, the jury could not have concluded that Demag misrepresented anything concerning the warranted qualities of the caster. [13] Copperweld contends that the following errors committed in the trial of its case require reversal: (1) that the trial court, over Copper-weld’s objection, permitted a Demag expert witness to testify about an expert study without a proper foundation for the testimony; (2) that the trial court, by sua sponte rulings, unduly restricted Copperweld’s rebuttal; (3) that the trial court allowed Demag to introduce evidence on collateral issues prejudicial to Cop-perweld; and (4) that the trial court’s response to jury questions posed during deliberations was inadequate. We find the district court’s actions within its discretion and reject these claims. [14] Although counsel for Copperweld obtained statements from Holmquist concerning his views on the casting machine, Demag did not itself have an opportunity to depose Holmquist prior to his death. The memorandum was an essential document, not protected by an attorney-client privilege and the district court acted within its discretion in ordering production of the document. Fed.R.Civ.P. Rule 26(b)(3) (Party seeking discovery has substantial need for materials which cannot be discovered without undue hardship); cf. Hickman v. Taylor, 329 U.S. 495, 509-14, 67 S.Ct. 385, 91 L.Ed. 451 (1947) (court affirms denial of production of statements of witnesses in defendant’s file where public record of testimony existed and no necessity for the attorney’s files demonstrated). [15] Rule 804(b)(5) of the Federal Rules of Evidence, codified at 28 U.S.C. (Supp. V 1975), provides in pertinent part: (b) Hearsay exceptions. — The following are not excluded by the hearsay rule if the de-clarant is unavailable as a witness: ****** (5) Other exceptions. — A statement not specifically covered by any [other exception] but having equivalent circumstantial guarantees of trustworthiness, if the court determines that (A) the statement is offered as evidence of a material fact; (B) the statement is more probative on the point for which it is offered than any other evidence which the proponent can procure through reasonable efforts; and (C) the general purposes of these rules and the interests of justice will best be served by admission of the statement into evidence. [16] The memorandum was prepared by one of Copperweld’s counsel, Richard C. Seamans. It was stipulated that it was unnecessary to call Mr. Seamans to confirm that he wrote the memorandum. Furthermore, the parties agreed that Mr. Holmquist had in fact edited the memorandum and that the handwritten notes on the document were his. [17] We believe that more probative evidence of the content of this memorandum could have been provided by Attorney Seamans. Because he was not called as a witness, the memorandum was in fact double hearsay — that of Mr. Holmquist as stated or interpreted by that of Mr. Seamans. Copperweld, however, has waived its objection to this problem by its stipulation that the document be admitted without the testimony of the scrivener. [18] The district court gave the following reasons for its decision not to enforce the forum selection clause: (1) that the facility which was the subject of the action was located in Warren, Ohio, and was likely to become the object of an intensive inspection during the trial; (2) that the facility was fabricated by Birdsboro Corporation, a Pennsylvania contractor, in this country; (3) that all of the records concerning operation of the plant were in this country; (4) that all of Copperweld’s personnel who operated the plant were in this country; (5) that all of Cop-perweld’s personnel who negotiated the contract were in this country; (6) that certain of Demag’s personnel involved in the sale were in this country and that Demag was doing business in the United States and maintained offices in Pittsburgh, Pennsylvania; (7) that practically all of the activities undertaken in connection with the contract took place in the English language; (8) that almost all of the witnesses were English speaking; and (9) that conducting the litigation in Germany would have required translation with inherent inaccuracies. Copperweld Steel Co. v. Demag-Man-nesmann-Boehler, 54 F.R.D. 539, 542 (W.D.Pa. 1972). Furthermore, all of the plant’s customers were in this country and if their testimony were necessary, the district court envisioned difficulties in compelling their attendance in a German forum. Copperweld Steel Co. v. De-mag-Mannesmann-Boehler, 347 F.Supp. 53, 54 (W.D.Pa.1972). [19] It is significant to note that the Supreme Court has cited the Central Contracting Company cases with approval. See The Bremen v. Zapata Off-Shore Co., 407 U.S. 1, 10 n. 11, 92 S.Ct. 1907, 32 L.Ed.2d 513 (1972).