Krause v. Sud-Aviation, Societe Nationale De Constructions Aeronautiques

CROAKE, District Judge. MEMORANDUM DECISION FINDINGS OF FACT AND CONCLUSIONS OF LAW These are suits in Admiralty based upon the Death on the High Seas Act (DOHSA), Title 46 U.S.C.A. § 761 et seq. Libelants in their representative capacity seek to recover damages for the wrongful death of George T. Stubbs (the pilot), Luke Hampton Wade, Jr. (a passenger), and Harold J. Nicol (a passenger) in an Alouette II helicopter, U. S. Registration No. N519, which crashed in the Gulf of Mexico off Lee-ville, Louisiana, on November 30, 1959. The suits were consolidated for trial and by agreement of counsel, the issue of liability was first tried separately pursuant to Rule 42(b), Fed.R.Civ.P. JURISDICTION In its post-trial memorandum respondent Sud-Aviation, Societe Nationale de Constructions Aeronautiques (SUD) challenges the Court’s subject matter jurisdiction by disputing application of the DOHSA. The same challenge is expressly set forth at page 6 of the pretrial order, it being SUD’s contention that the accident occurred in the territorial waters of Louisiana at a location not “more than - a marine league” 1 from the shore of that state. This precise question appears to have been argued and determined already. Judge Bartels in an earlier phase of the litigation observed: “The libel alleges that the accident occurred ‘approximately twenty (20) miles south of Leeville, Louisiana’ and the Exceptive Allegations and Exhibit A attached thereto establish the scene of the accident approximately five nautical miles from the Louisiana *516 coast, which is admittedly ‘beyond a marine' league from the shore’ of that State.”' 196 F.Supp. 856, 858 (EDNY 1961). Although SUD’s reply memorandum (pp. 10-11) implies that Judge Bartels did not specifically find the essential jurisdictional facts required, we fail to see how this can be so in light of his conclusion: “ * * * that the accident having occurred ‘on the high seas more than a marine league from the shore’ of Louisiana, the Federal Death on the High Seas Act is applicable and this Court has jurisdiction over the first cause of action thereunder.” 196 F. Supp. 856, 859. In this Court’s view the matter of jurisdiction was settled by Judge Bartels. Stoll v. Gottlieb, 305 U.S. 165, 59 S.Ct. 134, 83 L.Ed. 104 (1938). But if that decision is not res judicata or “law of the case,” the issue must turn on evidence presented at trial. The report of Jacques Corniot, an investigator employed by SUD, states that shortly after the accident the wreck of N519 was found floating upside down “approximately 12 miles from the shore.” 2 At page 45 of its memorandum SUD points to certain other vague references to the location of the accident. We think these references do not support the inferences SUD draws from them. Because Corniot’s report is apparently otherwise uncontradicted, the Court finds that the accident occurred more than a marine league from shore and that jurisdiction exists under the DOHSA. BACKGROUND N519 was built by respondent SUD and sold as a new aircraft to respondent Republic Aviation Corporation (Republic or RAC) pursuant to a supplemental agreement dated November 14, 1957. The aircraft was delivered in France in February 1959 and shipped to the United States in a partially disassembled state by Republic, who reassembled it and used it for approximately 437 flight hours as a demonstration model. Republic then leased it “as is” to Petroleum Helicopters, Inc. (PHI) for use in its business of supplying air transportation-to companies conducting oil exploration in the Gulf of Mexico. On the day of the crash, PHI pilot George Stubbs was ferrying Leslie H. Wade, Jr., and Harold J. Nicol, employees of Gulf Oil Corporation, to an off-shore oil rig in the Gulf of Mexico. Stubbs reported by radio that he thought he had experienced a “pitch-change, failure” 3 and that he- was making a 180 degree turn in order to return to. shore. There was some additional exchange by radio in which Stubbs discussed the difficulties with his supervisor, Joseph Bolen. . Shortly thereafter the people monitoring the radio heard a high-pitched scream. This, followed by loss of .radio contact, was the first sign of disaster. Other helicopters were immediately dispatched to the last-reported position given by Stubbs. The first' people to arrive at the scene of the accident found, several items, including á logbook, floating in the water but all persons aboard had perished. With benefit of hindsight it has been established that what Stubbs believed was a “pitch-change failure” was in reality a “failure” or break, in the structural steel of the helicopter’s tail. This tail 4 is constructed so that' standing directly behind it, the structure would appear as the cross section of' a prism with the apex of the triangle *517 pointed toward the ground. The edges of the prism are formed by three longerons of tubular steel which extend back from the main body. The longerons are connected to one another and supported by a web of steel supports running between them at various intervals along the structure of the tail boom. The entire tail assembly tapers as it moves aft so that the triangle formed by the three longerons is smaller at the rear, of the aircraft than near the front. Approximately two-thirds of the way back along this tail assembly is a horizontal stabilizer designed to control or stabilize the tail during flight. 5 It is basically an air foil mounted in the horizontal plane perpendicular to the prism described above. The stabilizer is mounted in two steel brackets which in turn are welded to the upper right and .upper left longerons, respectively. 6 It is not disputed that the cause of the. crash was the failure of the upper right longeron at the point where the horizontal stabilizer bracket was welded to it. 7 This was followed after several minutes by other failures in the structure of the tail boom. 8 ' Libelants allege negligence and breach of warranty against each respondent. Against SUD the charge is that the weld of the horizontal stabilizer bracket to the upper right longeron was negligently made, resulting in a condition known as “insufficient root penetration.” This in turn led to untoward, stresses near the point of failure. Republic is charged with not having properly inspected the aircraft for- defects after reassembling it in the Unitéd States. It is also asserted that both respondents impliedly warranted the fitness and merchantability of the aircraft and that these warranties were breached. Both respondents take the position that the allegations of negligence are unsubstantiated by the proof and that no warranties existed. Respondents affirmatively allege that pilot Stubbs was contributorily negligent and that intervening negligence on the part of PHI was the proximate cause of the crash. SUD’S NEGLIGENCE The principal allegation of negligence directed toward respondent SUD is that one of the welds joining the horizontal stabilizer bracket to the failed longeron was defective. The defect alleged is that of “insufficient root penetration” and much of the testimony was devoted to expert opinions on whether or not such a condition existed. The weld under consideration is basically T-shaped and the point at which the two surfaces are in' contact with one another is known as the “root” of the weld. 9 The experts were in substantial agreement that in a properly welded joint sufficient heat is applied during the welding operation to fuse, these two surfaces to one another as well as to the weld metal. If too little heat is applied, so that this fusion of the parent metals does not occur, the result is described as “insufficient root penetration.” 10 According to libelants’ expert, •Isaac Stewart, in a weld with adequate root penetration, stress applied to • one piece of metal is transferred directly to the other piece, that is," the two pieces are fused so as to transmit stress as if they were one. Where the root penetration is . inadequate, however, and the two parent pieces are not fused, stress transfers between them are sidetracked and pass through the weld metal. 11 The alleged adverse effects of this rerouting of stress are twofold. 12 First, *518 an increased amount of stress is introduced in the necessarily weak heat affected zone 13 around the weld. Second, a bending stress is introduced that would not otherwise be present. All things considered, Mr. Stewart’s explanation presents the most viable theory of how the particular break pictured in libelants’ exhibits 9, 11 and 12-A occurred. 14 Lending support to this “displaced stress” theory are the following considerations: 1) Stresses initiating in the horizontal stabilizer would have to be transferred through the bracket to the failed longeron along one of the routes described by Stewart. The coincidence of a change in the stabilizer just before the crash 15 leads almost inescapably to the conclusion that the source of those stresses that were the immediate cause of the failure was the new stabilizer. 2) The weld metal did not break. Rather, the longeron tube itself broke in a “Y” shape along 8 basic lines: one at the junction of the two welded pieces; another along the edge of the weld fillet; and the third at the side of the tube directly opposite the first two 16 The fact that one breakline ran along the edge of the bracket lends support to the theory that a bending stress was applied that the longeron was unable to withstand. It appears that the bracket may have acted as a fulcrum around which the longeron was bent when the displaced stress was applied at a point on the longeron some distance away from the root of the weld. "~ 3) The fact that the second breakline runs along the edge of the weld fillet also supports the theory that some of the stress was displaced through the weld metal to the weaker, heat-affected zone. Although these considerations lend support to an “insufficient root penetration” theory of the accident, they are by no means inconsistent with the theory that the weld was perfectly sound and the longeron merely subjected to extraordinary stresses which caused the failure to occur at a point weakened by a perfectly proper welding process. But the failure having occurred under normal flight conditions, the case for deficiency based on circumstantial evidence is bolstered by the very occurrence of the failure. That is, absent some intervening cause either the failed part (or the design of the aircraft) was necessarily defective. The establishment by libelants of such a prima facie case shifts the burden of going forward to the respondent SUD who must show that abnormal stresses rather than defective construction were the cause of the failure. 17 SUD attempted to meet this burden by introducing evidence of “rough handling,” including an incident in which the rear rotor of N519 was dipped into the water during a practice landing on August 19, 1959. That there was a tail-dipping incident is not in doubt, the disagreement is as to its magnitude *519 and effect on the failed longeron. Joseph R. Bolen, an experienced helicopter pilot and supervisor of the deceased pilot Stubbs, testified that instances of the tail rotor blades striking either land or water, although “serious,” were not uncommon especially where the technique of landing by autorotation was being practiced. 18 Bolen further testified that the effect of such an incident with respect to subjecting the tail boom of the helicopter to abnormal stresses would depend on the force of the strike. 19 He stated that “the incident we are referring to in August was not considered a hard strike. This is a strike that would happen on any helicopter in the course of normal training.” 20 Bolen’s testimony that the strike in question was not a hard strike receives support from the testimony of John Howren, the PHI mechanic who repaired the damage. Howren testified that he made an inspection of the moving parts involved in the tail rotor assembly which showed that the only damaged part was a six-inch section of one of the tail rotors. Howren also reported having inspected the failed boom, including the welds. 21 His observation as a result of all of these inspections was: “I saw no damage to any of these parts that were removed or to anything other than the one tail rotor blade as I described it, deformed and slightly dented, no cuts, abrasions, or anything wrong with the tail blade or any other, part of the helicopter.” 22 That the strike involved in this particular tail-dipping incident was not severe is further corroborated by the fact that the two pilots involved flew the plane back to base immediately after the accident. 23 They would hardly have done so if they thought there was any risk to themselves. Upon touchdown they apparently reported no abnormal vibration or other unusual effect upon the flight of the plane. There is, however, contrary evidence tending to show that the tail-dipping incident was relatively serious. First, notwithstanding Howren’s testimony that the only damage was to a single rotor blade, the entire tail rotor assembly including the drive shaft was replaced at a cost of more than 8 man hours and several thousand dollars in parts. 24 Second, two of respondents’ expert witnesses testified to the amount of energy that necessarily had to be dissipated in a very short period of time when a tail rotor of a certain length traveling at a certain r.p.m. is dipped into water. 25 As to the first point, the mechanic Howren testified on cross-examination that the additional replacements were precautionary and were made on the instruction of Howren’s supervisor at Lafayette so that the parts could be returned there, disassembled and further inspected, “and probably returned to service if there was nothing wrong with them.” 26 The second point, that a severe stress is necessarily created when a rotor blade strikes the water, has several weaknesses. First, it assumes that the rotor stopped completely or nearly so. Second, although undoubtedly torque stresses were created in the mechanism of the tail rotor assembly, the follow-up testimony attempting to show how these unusual stresses would be transferred to the structural frame of the aircraft and how severe such transferred stresses would be was not persuasive. 27 Third, this “necessarily severe stresses” testimony was theoretical in nature. To the extent that it implied there must be serious damage to an aircraft as a result of any tail-dipping incident, it was directly contradicted by the testi *520 mony of both the pilot Bolen and the mechanic Howren as to their experience with helicopters that had been subjected to tail rotor strikes. 27 Respondents’ remaining evidence of “rough handling” by PHI is almost exclusively a detailed list of repairs. 28 Counsel cites no evidence of specific instances of “rough handling” that led to these repairs, nor was there expert testimony that the nature and quantity of repairs was such that rough handling might be inferred. Under the circumstances, inferences that N519 was a peculiarly repair-prone aircraft, or that the PHI maintenance staff was exceptionally careful, are as consistent with the evidence as is an inference of “rough handling.” There is one final and overriding reason for rejecting the theory that the fracture of the longeron was caused by the tail-dipping incident and/or rough handling. This is the failure of respondents to demonstrate that the fracture was a gradual or “fatigue” fracture rather than a sudden or “brittle” fracture. The tail-dipping incident occurred on August 19, 1959 after some 462 hours of flight and the failure on November 30 after 622 hours. This means that there were approximately 160 hours of safe, normal flight over a three-month period between the tail-dipping incident and the crash. To sustain the burden of explanation respondents must show that the break began at the time of the tail-dipping incident or as a result of some other incident of rough handling and progressed by fatigue until the day of the crash. The evidence and testimony of the experts on this point were inconclusive. None of the experts observed the classic signs of a fatigue failure. Holshouser, who examined the parts soonest after the accident, stated that although a fatigue fracture was possible, his examination did not lead him to a definite conclusion as to whether the failure was progressive or not. 29 Stewart, libelants’ primary expert, never examined the parts but, rather, based his analysis on photographs of the parts and reports. Stewart testified that while the photograph marked libelants’ exhibit 12-A did not show the classic markings of fatigue, there were areas of the photograph that were not clear enough to interpret one way or the other. 30 Only Chapman, respondents’ expert, took a firm position on the question of whether the break was sudden or progressed by fatigue. Chapman stated that he “believed” there were “some fatigue propagations of the crack from the two ends of the ears of the bracket.” 31 The basis of this belief was apparently the fact that “there were some darkened areas out near the ears in the outboard thing.” 32 Chapman stated that he wanted to clear off these darkened areas and take a look, but because the parts were at the National Bureau of Standards for testing (by Holshouser) he could not do so. When Chapman finally did get the parts back to his laboratory, “they were so badly rusted that anything on that fracture face was obliterated.” 33 On the basis of his observations and testing of the parts, Chapman ultimately testified that in his opinion the force that caused the tube to fracture was “a longitudinal force applied along the upper right-hand longeron in a very, very small period of time, so short as *521 to constitute a hammer blow impact, very, very suddenly-applied blow.” 34 This created the initial break which progressed by fatigue to the final failure. Several factors lead the Court to reject Chapman’s explanation of the accident. First, late in the afternoon of the day preceding the above-described testimony, Chapman clearly and unequivocally testified that the crack he observed in August 1960 “was not a fatigue crack.” 35 It is hard to explain his testimony the following day as anything but a reversal or substantial modification of this earlier opinion. Second, Chapman described the force causing the break as a longitudinal force along the upper right longeron in the nature of “an impact blow putting the entire longeron in tension, * * *.” 36 Although this blow allegedly occurred “at the time of” the tail-dipping incident, the testimony of respondents’ witnesses was just not adequate to explain how dipping the plane’s tail rotor in the water would result in the impact and the tension described. 37 Finally, as discussed earlier, most of the evidence points to this tail-dipping as having been relatively mild. If the tail-dipping was the source of the failed longeron in this instance, it would seem possible to produce evidence that tail strikes are causally associated with structural failure. No such evidence was introduced. In sum, we reject the theory that the fracture began at the time of the tail-dipping incident and progressed by fatigue until the day of the crash. In line with our holding, supra, that libelants’ prima facie case shifts to respondent SUD the burden of establishing as a cause of the accident some factor unrelated to a defective part, we find that SUD has failed to meet that burden and is liable in negligence. CONTRIBUTORY NEGLIGENCE OF PILOT Respondents have argued that libelant Stubbs was negligent in not descending immediately and landing on the water. If Stubbs is found to have been negligent, there would be no effect on the ability of the other two libelants to recover 38 and even as to Stubbs a comparative negligence standard would apply. 39 It is easy, with benefit of hindsight, to say what Stubbs should or should not *522 have done. But at the moment in question, Stubbs had neither the information presently available nor the detachment that comes from viewing the problems ex post and from the ground. Stubbs incorrectly diagnosed his difficulty. He assumed from the fact that the plane lurched to the right that he had experienced a pitch control failure. Respondents argue that if Stubbs had been sufficiently knowledgeable of the aircraft he was flying, he would have realized that because the direction of ro-' tation of the Alouette’s tail rotor is opposite to that of most American helicopters, a lurch to the right could not possibly signify a pitch change failure. When the problem is thought through carefully, 40 this analysis seems correct. But there was no testimony to the effect that the symptoms experienced by Stubbs would only be consistent with a broken longeron or with the kind of emergency that would call for an immediate landing by autorotation. Indeed, Stubbs was in radio contact with two other experienced helicopter pilots, neither of whom suggested that procedure. Stubbs experienced certain symptoms and like a doctor or any other specialist operating under pressure of time and circumstance he made an educated guess. The fact that the guess ultimately proved wrong or even disastrous does not make it negligent. See, e. g., Kelley v. Safeway Stores, Inc., 105 U.S.App.D.C. 406, 267 F.2d 683 (1959); Prosser, Torts, § 33, n. 82 (1964 ed.). The Court finds that under the circumstances Stubbs was not negligent. RAC’s NEGLIGENCE Libelants allege that respondent Republic was negligent in not properly inspecting the aircraft when it was assembled and/or when.it was leased “as is” to PHI. Assuming arguendo that RAC had a duty to make such inspection, in light of the Court’s finding that, whatever its cause, the break in the longeron was not gradual, it follows that there could be no causal connection between RAC’s failure to inspect and the break in the longeron. Even if RAC had conducted the inspections it allegedly failed to perform, at that time there would have been no discoverable defect. 41 It follows that respondent Republic’s negligence, if any, was not causally related to the failure of the longeron. PHI’S NEGLIGENCE Respondent SUD also makes the point that another, and perhaps more significant inspection, was not conducted by PHI. This inspection, required by factory maintenance standards to be conducted after 600 hours of flight, would have included examination of the tail boom for cracks. Apparently as the result of an error on the part of the PHI maintenance staff, when the helicopter was leased from RAC, a new maintenance and inspection cycle was begun. The result was to advance the schedule of services conducted at 50-hour intervals, but to retard the special additional inspection and servicing conducted at only 150 and 300-hour intervals. At the time of the crash, the aircraft had flown 622 hours, or 22 hours past the time for inspection of the tail boom. The argument that PHI’s negligence in failing to inspect was a proximate cause of the crash assumes that the defect was there to be found at the time of the scheduled inspection. Having rejected the tail-dipping-fatigue hy *523 pothesis, it follows that the defect was not discoverable as of 600 hours and the negligence of PHI, if any, in failing to inspect could not have been a proximate cause of the crash. IMPLIED WARRANTY Having found that respondent SUD was negligent, that respondent RAC was not, that there was no contributory negligence on the part of Pilot Stubbs, and no intervening negligence on the part of PHI, we turn now to the question of warranty. It is settled that under the DOHSA the applicable law is federal admiralty law. E. g., Middleton v. Luckenbach SS. Co., Inc., 70 F.2d 326 (2d Cir.), cert. denied, 293 U.S. 577, 55 S.Ct. 89, 79 L.Ed. 674 (1934); D’Aleman v. Pan American World Airways, Inc., 259 F.2d 493, 495-496 (2d Cir. 1958). In simplified form, then, the question is whether the doctrine of implied warranty is applicable “in admiralty.” Three judges in this district have held that it is and there are other cases supporting that view. Middlleton v. United Aircraft Corp., 204 F.Supp. 856 (SD NY1960); Sevits v. McKiernan-Terry Corp., 264 F.Supp. 810 (SDNY1966); Montgomery v. Goodyear Tire and Rubber Co., 231 F.Supp. 447 (SDNY1964). There is also an extensive decision by Judge Layton of the District of Delaware taking the opposite position. Noel v. United Aircraft Corp., 204 F.Supp. 929 (D.Del.1962). See also Jennings v. Goodyear Aircraft Corp., 227 F.Supp. 246 (D.Del.1964). After its thorough and careful analysis, the Court in Noel held, “that admiralty (even though it could) should not at this time adopt the implied warranty of fitness and merchantability doctrine into the federal maritime law.” 204 F.Supp. 929, 941 (emphasis supplied). In Judge Layton’s view, “[t]he balance of considerations [was] tipped towards denial of the Libellants’ motion [to amend their complaint to include an implied warranty theory] by the degree of change the Court [was] asked to make in the law, under the facts [there] presented.” 204 F.Supp. 929, 939. The undersigned follows the reasoning of the Noel court and, indeed, might well have reached the same result in 1962. But in the six years between that decision and this, the situation has changed sufficiently to allow the same reasoning to lead to the opposite conclusion. 42 The doctrine of implied warranty (in various forms) has now achieved wide acceptance as a theory of recovery in air disasters occurring over land. 43 In the *524 absence of international complications, 44 we see no reason for a different standard to apply merely because the plane happened to be operating over water. The DOHSA is primarily a statutory authorization of a right of action “in admiralty” with the substance of that right free to grow and change with federal maritime law so long as there is no conflict with the basic statutory purpose. The purposes for which a statute is passed and the assumptions of its drafters are not identical. Notwithstanding the fact that the Congress which passed the DOHSA probably assumed that a negligence standard would apply, the purpose of the statute was to grant a “right of action” that would remain flexible. If this view is correct, absorption into the law applicable under the DOHSA of the liability standards implicit in implied warranty is appropriate. This Court therefore follows the decisions in Middlleton v. United Aircraft and Sevits v. McKiernan-Terry Corp., supra, and holds that, as a general matter, nothing precludes the application of implied warranty to DOHSA cases. Although breach of implied warranty is in one sense a form of absolute liability, or liability without fault, in another sense it is not. The “fault” involved is breach of the warranty, but this is in a way tautological because the warranty is always “implied” ex post to fit the facts. Nevertheless, we cannot and do not assign liability at random or on a “deep pocket” basis; a true “deep pocket” theory would simply make all injury, whatever the source, compensable by some giant central insurer. Until that nirvana is achieved, we are faced with the task of rationalizing who warrants what to whom and why. A warranty is in essence a promise. We think it is reasonable for a passenger in an aircraft to proceed on the assumption that the manufacturer has made the plane, if not perfectly, at least in such a way that its ordinary use will not prove harmful. The very fact of manufacturing and offering the airplane for sale implies a promise to this effect made to all persons who might reasonably be expected to ride in it. 45 Accordingly, we hold that respondent SUD impliedly warranted that its helicopter would not prove dangerous or harmful in ordinary use. We further hold that by virtue of the crash on November 30, 1959 that warranty was breached. 46 Whether Republic made any implied warranties is a more difficult ques *525 tion. The primary basis for Republic’s liability in implied warranty is its ownership of the aircraft. This ownership carried with it no element of control, however, and there is no allegation that decedents relied on Republic’s name or reputation or, indeed, that they even knew that Republic owned the aircraft. In the absence of some relationship other than mere formal ownership, we see no reason to hold that Republic impliedly warranted anything to these libelants. In essence, Republic was nothing more than the first owner and operator of a second hand or used helicopter that was leased rather than sold to PHI. As such, like any interim owner, Republic may have owed certain duties of care and inspection, breaches of which would sustain liability in negligence. But any warranty would be separate and distinct from these duties and we hold that no such warranties were made or breached. RELEASE OF PHI One final matter remains to be disposed of. In 1960 plaintiffs Wade and Nicol brought actions, separate from the present one, in Louisiana District Court against these same defendants plus PHI. The actions against PHI were settled and apparently releases were executed. In its post-trial memorandum SUD argues that these releases running to PHI also released defendants SUD and RAC. SUD’s argument rests on a doctrine of Louisiana law which purportedly holds that when several parties are sued in solido, unless rights are expressly reserved against the parties not intended to be released, a release of one party is a release of all. As Exhibits “A” and “B” attached to their post-trial memorandum, plaintiffs have submitted xerox copies of releases running to PHI from Mrs. Nicol and from the five Nicol children. Each of these releases contains the following paragraph: It is specifically understood that [the party executing the release] is releasing hereby only [his] claim against the above named party, its agents, employees, heirs, successors, insurers and assigns, and that said [party executing the release] does hereby expressly reserve any claims, right, or cause of action that [he] may have against Republic Aviation Corporation and/or Sud-Aviation and does hereby expressly reserve any right that [he] may have to proceed against any or all of said parties. Neither of these documents was offered in evidence and no similar document covering the Wade claim was submitted. Although the issue was not mentioned at trial, plaintiffs cannot plead surprise because the issue was specifically preserved at page 7 of the pre-trial order as item No. 7 in the list of defendant Republic’s contentions. The fact of the Louisiana claims having been brought and settled was read into the record at trial, 47 although there was no specific proof of a release having been executed. Whether all this is enough to satisfy the affirmative defense requirements of Rule 8(c), Fed.R.Civ.P., we do not have to decide. It appears that the doctrine on which SUD relies has been authoritatively construed to require that the party asserting it show that the party to whom the releases run was in fact a joint tortfeasor. Reed v. Rheam Manufacturing Co., 364 F.2d 810 (5th Cir. 1966). We have already found that the pilot was not contributorily negligent and that the intervening negligence, if any, on the part of PHI was not a proximate cause of the crash. It follows from these findings that SUD has not shown that PHI was a joint tortfeasor and that the doctrine upon which it relies is inapplicable. This establishes that the release of PHI by the Wade and Nicol plaintiffs did not release SUD or Republic. This is not to say, however, that the amount received in settlement of those claims *526 may not be treated as mitigating the damages suffered by the plaintiffs. CONCLUSION Let a decree in accordance with the foregoing Findings of Fact and Conclusions of Law be settled before the undersigned. On the return date of said notice of settlement the attorneys may be heard on the question of damages. So ordered. Footnotes [1] One .marine league is equal to three nautical miles of 6080.20 feet each. [2] SUD’s Ex. 13-A, p. 2. [3] The rear rotor of an Alouette II helicopter is designed to offset the torque generated by the main rotor and prevent the helicopter from spinning in the air. To control the amount of anti-torque applied it is possible to adjust the “pitch” or angle of attack of the tail rotor blades from_ the cockpit of the aircraft. It was this ability that pilot Stubbs assumed he had lost. [4] See libelants’ exhibits 1-A, 1-B, 17 & 18. [5] See libelants’ ex. 17. [6] See libelants’ ex. 18, SUD’s ex. G. [7] See libelants’ ex. 18, SUD’s ex. G. [8] See libelants’ ex. 9. [9] Tr. 547, 550. (References preceded by “Tr.” are to the trial transcript.) [10] Tr. 704-5. If too much heat is applied an equally undesirable condition known as “burn through” results. [11] See libelants’ ex. 15, figure 2; tr. 545-54. [12] Tr. 553. [13] Apparently even a perfectly made weld weakens the parent metal to some extent. See tr. 536-8. [14] See also SUD’s ex. G. [15] Tr. 242-9. [16] See libelants’ ex. 15. [17] Cf., e. g., Lobel v. American Airlines, Inc., 192 F.2d 217 (2d Cir.), cert. denied, 342 U.S. 945, 72 S.Ct. 558, 96 L.Ed. 703 (1952); Rogow v. United States, 173 F. Supp. 547, 556 (S.D.N.Y.1959) (Kaufman, J.). The doctrine of res ipsa loquitur applied in these two cases does not apply here and indeed plaintiffs disclaim reliance on it in the pre-trial order. Nevertheless, the basic principles of circumstantial proof on which the doctrine of res ipsa rests are applicable. There is no doubt that the failure in this instance was in the structural steel of the aircraft. Whether this fact alone would be enough to shift the burden of explanation to the manufacturer we do not have to decide. But certainly the fact of such failure coupled with (a) proof that it occurred under normal flight conditions and (b) plausible expert testimony that the structure was defective when made is enough to shift that burden and we so find. [18] Tr. 135-7. [19] Tr. 135-6. [20] Tr. 136. [21] Tr. 228-36. [22] Tr. 234. [23] Tr. 222-3. [24] Tr. 235-6; 386-7. [25] Tr. 772-5 ; 982-3. [26] Tr. 387. [27] See footnote 37 at p. 521, infra. [27] Tr. 392; 136-7. [28] SUD’s brief, pp. 10-12. [29] Tr. 494. Objection was made to the admission of Holshouser’s testimony. While the objection may have some validity with respect to Holshouser’s opinions as an expert, the Court does not rely on any such opinions. We think the objection is not well taken with respect to factual observations. E. g., Berguido v. Eastern Airlines, Inc., 317 F.2d 628, 632 (3d Cir. 1963). [30] Tr. 651. [31] Tr. 787. [32] Tr. 787. [33] Tr. 787. [34] Tr. 787. [35] On direct examination Chapman testified : Q Mr. Chapman, with respect to the crack, the fracture in a tube which you observed in August 1960, do you have an opinion as to whether that fracture was the result of fatigue operating from the date of manufacture of the helicopter in the absence of any unusual or specially severe force being applied to the tail section? A Yes. Q What is your opinion? A That it is not a fatigue crack. Q What is the basis for that opinion? A I cannot find the characteristic markings on it. (Tr. 772.) [36] Tr. 803. [37] The testimony based on mathematical calculations as to the force that necessarily must be dissipated in a relatively short time, if directed to explain the source of the stress described, is madequato. The stress generated presumably would be relayed mechanically in a number of directions through the various linkages of the tail rotor drive as testified to by RAO’s expert, William Cobey. (Tr. 982-3.) To the extent that that stress was eventually transmitted to the longerons, it would have been divided equally among all three of them. We think the nature and extent of any force ultimately exerted upon the upper right longeron as a result of the tail-dipping incident is not a matter that can be assumed or left to conjecture. The attempts of Cobey to supply adequate proof in this area simply were not persuasive. In the course of a direct confrontation on cross-examination Cobey was at worst evasive and at best unable to communicate how and to what extent the tail-dipping would have affected the failed longeron. (See Tr. 1016-24.) [38] See generally Prosser, Torts, § 73 (1964). [39] 46 U.S.C.A. § 766. [40] The complexity of respondents’ explanation (SUD’s Brief, pp. 16-17) shows just how carefully the problem must be thought through. It is not surprising that under the circumstances Stubbs did not go through the same analysis. [41] This is true even under respondent SUD’s theory that the break commenced with the tail-dipping incident and progressed by fatigue until the ultimate failure. Any theory that envisions a discoverable defect in existence during the time RAO was operating the helicopter is unsustainable in light of all the evidence. [42] Since the decision in Noel, one district court has expressly declined to follow it in a DOHSA action, Sevits v. McKiernan-Terry Corporation, 264 F.Supp. 810 (SDNY 1966), and one other decision, although it does not discuss Noel, expressly allows an implied warranty action under the DOHSA, Montgomery v. Goodyear Tire & Rubber Co., 231 F.Supp. 447 (SDNY 1964). Also decided since Noel are the following cases which indicate acceptance of the doctrine of implied warranty in airplane disasters where state law was being applied. Goldberg v. Kollsman Instruments Co., 12 N.Y.2d 432, 240 N.Y.S. 2d 592, 191 N.E.2d 81 (Ct.of Apps. 1963) (Manufacturer held liable; privity not required); King v. Douglas Aircraft Co., 159 So.2d 108 (Fla.App.1963) (Followed Kollsman Instruments, supra, reversing summary judgment for defendants); Boeing Airplane Co. v. O’Malley, 329 F.2d 585 (8th Cir. 1964) (Helicopter crash; Pennsylvania law applied; language in contract was insufficient to disclaim implied warranty); Banko v. Continental Motors Corp., 251 F.Supp. 229 (B.D.Va. 1966) (implied warranty recognized as a theory of recovery, although proof failed; warranty defined as “an assurance that the product is free of latent defect, * * * and fit for the purpose for which it was made.”); Swain v. Boeing Airplane Co., 337 F.2d 940 (2d Cir. 1964) (defining certain elements that must be shown before recovery will be allowed under a warranty theory). [43] In addition to the post-Noel cases cited in footnote 42, supra, see Siegel v. Braniff Airways, Inc., 204 F.Supp. 861 (SD NY 1960); Hinton v. Republic Aviation Corp., 180 F.Supp. 31 (SDNY 1959); Conlon v. Republic Aviation Corp., 204 F.Supp. 865 (SDNY 1960); Ewing v. Lockheed Aircraft Corp., 202 F.Supp. 216 *524 (D.Minn.1962); Garon v. Lockheed, 7 Av. Cases 17,418 (Cal.Super.Ct., La. County 1961). [44] There is one very important factor that distinguishes Noel from the present case. Noel involved an accident to which the Warsaw Convention applied. Allowing suits directly against manufacturers effectively circumvents the strict damage limitations set forth in that document. As Judge Layton rightly points out, 204 F.Supp. 929, 937 n. 15, whether or not one agrees with the policies expressed in the Warsaw Convention, they are nevertheless a clear expression of Congress affecting matters that may arise within the admiralty jurisdiction of the federal courts. That expression of policy must, of course, be given weight in those cases where it applies. We do not face the problem in this case and take no position on how the question should be resolved. [45] Much has been said on the subject of whether implied warranty actions sound primarily in tort or in contract, and more particularly on whether or not privity of contract is essential to the maintenance of such actions. See Prosser, Torts, § 97, pp. 672-85 (1964), especially n. 41 and items there cited. See also Frumer & Friedman, Products Liability, § 16 (1967). Without reviewing these discussions, suffice it to say that we think the presence or absence of privity, although a factor to be considered should not be dispositive. See, e. g., Noel v. United Aircraft Corp., 204 F.Supp. 929, 935 at n. 12. [46] Implicit in this holding, of course, is the earlier finding of fact that the crash occurred during normal flight as the result of a structural failure the cause of which cannot be assigned to any person other than SUD. See Swain v. Boeing Airplane Co., 337 F.2d 940 (2d Cir. 1964). [47] Tr. 948-53.