Full opinion text
OPINION TENNEY, District Judge. Shortly after midnight on June 2, 1973, the container vessel SEA WITCH, owned by American Export Lines, Inc. (“AEIL”), was proceeding out of New York Harbor at full maneuvering speed when she lost control of her steering and veered to starboard, colliding after some four minutes or more, at substantially full speed, with the tanker ESSO BRUSSELS anchored in the federal anchorage off Stapleton, Staten Island, New York. Both vessels burst into flames; sixteen men, including the masters of both vessels, lost their lives, and many more were injured. Property damage was in the millions. Statement of Proceedings On June 7, 1973, AEIL filed a petition seeking exoneration from or limitation of liability under 46 U.S.C. §§ 183, et seq. (1982). Subsequently, numerous claims were filed by personal injury, death, cargo and other claimants against AEIL, which then counterclaimed against the ESSO BRUSSELS’ owner-operator and time and voyage charterers; Bond Hydraulic Equipment Service, Inc. and its successors (“Bond Hydraulic”); Sperry Corporation (“Sperry”); Bath Iron Works Corporation (“Bath”); and the United States of America (“U.S.A.”). On or about December 4, 1973, AEIL served a third-party complaint upon Bath seeking damages and indemnity within the limitation proceeding. This third-party complaint sets forth two causes of action: one in contract and one in tort. All of the aforementioned defendants cross-claimed against each other and against AEIL. However, although virtually all of the claimants were granted leave by the Court to file cross-claims against Bath, Sperry and Bond Hydraulic, many of them failed to ever do so. On January 25, 1977, the Court ruled that factual findings and exhibits contained in the Coast Guard Marine Board of Inquiry (“MBI”) record and report would be admissible, the report of the National Transportation Safety Board would be inadmissible, and the admissibility of testimony taken before the MBI would be determined at trial. See 73 F.R.D. 454 (S.D.N.Y.1977). Later in 1977, the Court granted requests by AEIL, Bath and Sperry for leave to assert cross-claims against the United States for contract and tort indemnity, and these were subsequently filed. A similar request on behalf of the ESSO BRUSSELS claimants was denied. See 76 F.R.D. 210 (S.D.N.Y.1977). On April 10, 1974, the United States filed cross-claims against Bath, Sperry and Bond Hydraulic, and they in turn reasserted claims over against each other in respect of the government’s claims. On July 21, 1983, the Court denied a motion by the United States for summary judgment dismissing the cross-claims of AEIL, Bath, and Sperry against it. See 568 F.Supp. 956 (S.D.N.Y.1983). AEIL contends that certain parties have discontinued their actions in exchange for assignments of their causes of action to AEIL, Pre-Trial Order (“PTO”) at 2-3, and that others have been dismissed or have discontinued their actions in exchange for a general release without assigning their causes of action. Id. at 3-4. Furthermore, AEIL has settled its third-party complaint against Bond Hydraulics, Sperry and the U.S.A. Id. at 4. Bath has reserved its rights with respect to any such settlements, releases or assignments, including any and all defenses which Bath may assert. Id. The remaining dispute in this action, and the sole issue before the Court, is AEIL’s claim for relief against Bath. The issue of liability was tried to the Court from May 7 through June 6, 1984. Thereafter, the parties filed voluminous briefs. AEIL essentially asserts that Bath, as manufacturer of the SEA WITCH and of the steering gear that failed, should be held liable for damages resulting from the collision. AEIL asserts three theories on which it should recover against Bath: negligence, strict liability, and implied warranties. AEIL in effect seeks both to recover for the direct damages it suffered and to be indemnified for sums paid to third parties in connection with the collision. Bath denies any and all liability with respect to the allegation that the steering mechanism was defective. Even though AEIL has settled with Bond Hydraulics, Sperry and the U.S.A., they remain parties to this action for the purposes of allocation of fault and any cross-claims. PTO at 4. AEIL denies any and all liability on the part of itself, Bond, the U.S.A. and Sperry. Id. Further, it is agreed among all parties that Bath shall not be liable to AEIL for any fault, liability or judgment attributed to anyone else. The Court finds that Bath is not liable in this action, and that the complaint against Bath should therefore be dismissed. The Court further finds that AEIL’s navigation of the SEA WITCH on the night in question was negligent, and contributed to the collision. Based on AEIL’s negligent navigation alone, the Court finds that 60% of the fault with respect to the collision must be allocated to AEIL. The Court concludes that AEIL may not recover from Bath for any or all of the remaining 40%. The Court has considered all of AEIL’s products liability claims, and finds that the evidence adduced at trial does not support those claims. Having given careful thought to all of the evidence and having weighed the credibility of all witnesses, the Court has set forth below its findings of fact and conclusions of law dealing with (1) the navigation of the SEA WITCH on the night of June 1-2, 1973, and (2) the design, installation, maintenance and operation of the steering mechanism aboard that vessel. Incorporated in these findings are the many stipulated facts which are germane to the disposition of the matters before the Court. NAVIGATION FINDINGS OF FACT 1. Shortly before midnight on June 1, 1973 at 2329, the container vessel SEA WITCH departed Howland Hook Container Terminal on Staten Island, New York, and proceeded to sea at slow ahead (4.7 knots) around the northern end of Staten Island by way of Kill van Kull under the control of a docking pilot with two Moran Company tugs in attendance. The vessel at that time was owned by AEIL. When the vessel reached a point near New Brighton, Staten Island, the harbor pilot, Captain John T. Cahill (“Cahill”), relieved the docking pilot who transferred to one of the two departing tugs. Trial Transcript (“Tr.”) 41-42; Exhs. 9, 2A. 2. SEA WITCH was at all relevant times an American flag vessel, home port New York, New York, 17,902 gross tons, 12,898 net tons, and approximately 594.2' long x 78.2' beam. At the times hereinafter mentioned she carried a crew of forty, including the master. PTO/UF 2. 3. As SEA WITCH proceeded down New York harbor, present on the bridge, in addition to the harbor pilot Cahill, were her master, Captain Paterson, who was in command of the ship; third mate Lovsin; deck cadet O’Connor; and the helmsman, able bodied seaman Louis Miller. PTO/UF 5. 4. At approximately 0025 on June 2, 1973, when SEA WITCH was off St. George, Staten Island, and clear of Kill van Kull, her speed was increased to half ahead, or about 9 knots through the water. The visibility was unlimited, the harbor was calm, and there was an ebb current of 2-3 knots. PTO/UF 6; Tr. 36; Exh. 2(a). 5. After passing St. George, those in charge of navigating SEA WITCH observed four vessels, including the tanker ESSO BRUSSELS, anchored in the federal anchorage 24, Stapleton, along the eastern shore of Staten Island. A tug and tow were observed southbound in the channel toward the Narrows, and a second tug and tow were observed leaving the anchorage and turning south toward the Narrows behind the first tug and tow. PTO/UF 7. 6. ESSO BRUSSELS was, at all relevant times, a tanker of 19,782 net tons, 667.4' long and with a beam of 97.3'. On June 1-2, 1973, she was loaded with 319,-402 U.S. barrels of crude petroleum and was anchored in position 40° 36' 46" North, 74° 03' 17" West, with her port side approximately parallel to the Staten Island shore, in proximity to the quarantine station. PTO/UF 1. 7. At 0029 on June 2nd, SEA WITCH’s speed was increased to harbor full, which was about 13V2 knots through the water. PTO/UF 8. 8. At about 0036, SEA WITCH passed Buoy 22 at a distance of about 700 to 1000 feet to port on a course of 167° True, which placed her about in the center of the channel heading approximately toward the center of the Verrazano Bridge. PTO/UF 9; Tr. 441. 9. A moment later the pilot, Cahill, ordered the course changed left, to 158° True to pass the two southbound tugs, KATHLEEN TURECAMO and BARBARA MORAN, on the starboard side of SEA WITCH. PTO/UF 10; Tr. 443; Exhs. 327, 328, 10, 12. 10. After executing the order to come left to 158°, to check the left swing, the helmsman applied 12° right rudder. At this point, steering control from the bridge was lost and the rudder remained locked at 12° right. At 0037.62, the helmsman reported that the ship was not steering. Tr. 438-45; Exh. 327. Upon receiving this report Captain Paterson exclaimed, “That damn steering again.” At this time, the distance between SEA WITCH and ESSO BRUSSELS was at least one mile. Tr. 129-30. 11. Upon hearing this report, the pilot, who was with Captain Paterson on the starboard side of the bridge, went to the center of the bridge; he noticed that the ship was heading straight down the Narrows, not toward Staten Island and not toward Brooklyn, and that the vessel’s head was drifting slowly to the right. PTO/UF 12; Tr. 52. 12. Captain Paterson proceeded to the steering stand and attempted unsuccessfully to regain steering by switching from the starboard steering system to the port steering system. PTO/UF 13. 13. Some time later, at about 0040, SEA WITCH passed about 150' ahead of the tug BARBARA MORAN and her tow. Thereafter, Cahill commenced sounding a series of short blasts on the whistle and then locked the whistle down to sound continuously. PTO/UF 14; Exh. 12, at 15-16. 14. When those in charge of navigating SEA WITCH concluded that collision was a certainty, Cahill ordered full astern. The third mate rang full astern on the engine order telegraph and the order was answered by the engine room. Thereafter, Cahill ordered the port anchor let go and ordered the general alarm sounded. PTO/UF 15. 15. SEA WITCH personnel stationed at the bow were unable to drop the port anchor because the riding pawl had not been raised prior to departure and was jammed against the anchor chain. The chief mate, on his own initiative, attempted to drop the starboard anchor; the anchor landed, however, on the deck of the ESSO BRUSSELS at collision. PTO/UF 16. 16. Seconds before collision, Cahill told all hands to clear the bridge and the bow. He left the ship’s whistle locked on and ran toward the stern of SEA WITCH. The bow of SEA WITCH collided with the starboard side of ESSO BRUSSELS at about 0042. The time between the helmsman’s report that steering was lost and collision was over four minutes. Tr. 446; Exhs. 327, 328, 329. Both vessels were immediately engulfed in flames and 16 men lost their lives. Property damage was extensive. PTO/UF 17. 17. Tests conducted by the Coast Guard and others on June 15, 1973 on STAG HOUND, a sister ship of SEA WITCH, brought the STAG HOUND from harbor full speed to a dead stop in the water at a distance of 520 yards. PTO/UF 17(A). 18. The Court finds that the distance between the SEA WITCH and the ESSO BRUSSELS when steering was lost was at least one mile and might have been as much as 1.3 miles. Tr. 129-30. 19. The time between the helmsman’s report that steering was lost and collision was about 472 minutes. The Court rejects Pilot Cahill’s contention that the interval was only 2 or 2V2 minutes. 20. The SEA WITCH had experienced prior steering difficulties and had recently undergone repairs to her steering system. PTO/UF 65. Those repairs were considered temporary. Id. Although AEIL knew the repairs were temporary, SEA WITCH’s after steering station was unmanned. The Court finds this to be contrary to the requirements of good seamanship. 21. Prior to the casualty, Captain Paterson had not advised Pilot Cahill of those earlier steering casualties, as he was required to do by common prudence and the practice of seamen. Pilot Cahill first heard of it when Paterson exclaimed “That damn steering again” upon receiving the helmsman’s report that steering was lost. Tr. 99. 22. The first and only engine order given by Pilot Cahill after the loss of steering and prior to collision was full astern. The Court finds that no stop order was given. Cahill’s testimony to the contrary is rejected. 23. Cahill waited about 3V2 or 4 minutes before ordering full astern, contrary to requirements of good seamanship, and even then he did not order emergency full astern, which would have called for more power. Tr. 89, 449, 507, 596. By this time, as Cahill admitted, collision was a certainty. Tr. 119. The SEA WITCH struck the starboard side of the anchored ESSO BRUSSELS at substantially full speed. 24. The Court finds that the failure to release the port anchor was also a factor contributing to the collision. Pilot Cahill testified that he hoped the anchor would “pull the bow to port,” away from the ESSO BRUSSELS. “It might have helped and it certainly couldn’t have hurt,” he concluded. Tr. 107. A further factor was the failure to drop the starboard anchor until the SEA WITCH was about to collide with the ESSO BRUSSELS. 25. In addition, the speed of the SEA WITCH down New York Harbor, about I6V2 knots over the ground, was excessive under the circumstances, and was not in keeping with the requirements of good seamanship. 26. Almost any timely action by Cahill would have avoided collision. Professor Norman A. Hamlin, whose testimony the Court fully accepts, demonstrated that by promptly stopping or reversing the engines, or dropping an anchor, the SEA WITCH would have stayed well clear of the ESSO BRUSSELS. Tr. 447-448; Exhs. 327-333. The Court rejects the testimony of AEIL’s expert Robert F. Stanley that no avoiding action was possible. THE STEERING GEAR ON THE SEA WITCH 27. The steering gear room on SEA WITCH contained, among other things, the steering engine, consisting of two hydraulic steering rams connected to the rudder shaft, the hydraulic steering pumps, the differential gear-box (or “differential”), the trick wheel, the connecting linkage, the rotary power units (“RPUs”), the tiller block, all related mechanical and hydraulic components, and the rudder angle transmitter. The steering gear room was located in the stern of the ship, just above the rudder. PTO/UF 46. 28. The steering engine pumps were activated by the differential which converted rotary mechanical signals into linear directional signals. PTO/UF 48. 29. The differential could receive mechanical rudder order steering signals from either the bridge, through two independently wired, hydraulically driven RPUs which converted electrical steering commands signals into rotary mechanical signals, or from a trick wheel mounted on the differential. PTO/UF 49. 30. The two RPUs were mounted on a single platform which in turn was mounted on the starboard end of the forward hydraulic ram cylinder. The RPUs were mounted next to each other, at the same height, with their output shafts facing aft. PTO/UF 51. 31. The differential was mounted opposite and above the RPUs on a separate platform which was mounted on the starboard aft ram cylinder. PTO/UF 52. 32. The differential input shaft was approximately eight inches higher than the RPU output shafts. The differential input shaft was directly aft of the port RPU output shaft by approximately thirty-five inches. PTO/UF 53. 33. The RPUs were connected to each other by a steel chain mounted on sprockets on the output shaft of each RPU. PTO/UF 54. 34.When the ship was being steered by use of the port RPU, the output shaft of the starboard RPU would be freewheeling because it would be rotated by the chain and sprocket. When the ship was being steered through the starboard RPU, the output shaft of the port RPU would be rotated by the chain and sprocket. If one RPU could not be operated, all steering commands could be transmitted by the other RPU. Only one RPU was needed to transmit the steering commands and only one operated at any one time. PTO/UF 55. CONCLUSIONS OF LAW In a collision case the court must look to the navigation or seamanship of the vessels involved. In the instant case, the ESSO BRUSSELS was properly at anchor in an official anchorage and was in no way at fault. Thus, it is the navigation and seamanship of SEA WITCH with which the Court is concerned. Bath argues that the seamanship and navigation of SEA WITCH on the night of June 1-2, 1973 was such that Bath should be exonerated from all liability, regardless of whether or not Bath was responsible for the steering failure. AEIL argues that the master and harbor pilot — AEIL’s agents— acted properly under the circumstances, and that the Court should not “second guess” an experienced master and harbor pilot or substitute its own judgment long after the event. In addition, AEIL contends that the loss of steering was the sole cause of the collision. For the reasons set forth below, the Court concludes that AEIL was negligent in its navigation and seamanship. The loss of steering and AEIL’s negligence were both causative factors in the ultimate collision. Based on a careful review of the evidence, the Court is convinced that SEA WITCH could have avoided the collision completely, despite the loss of steering, if proper steps had been taken. The Court further concludes that 60% of the fault of the collision must be allocated to AEIL for its negligent navigation and seamanship. A. AEIL’S NEGLIGENCE The faults charged against AEIL fall generally into two categories: precautionary and remedial. AEIL was negligent (1) in failing to take certain precautionary actions, prior to the loss of steering, that would have helped prevent a collision, and (2)in failing to take the remedial action needed to avoid collision once the steering was lost. 1. Precautionary Actions Those faults which may be described as precautionary are: (a) Failure of the master to advise the harbor pilot of prior failures of the steering machinery, so that appropriate precautions could be taken by the harbor pilot; (b) Failure to retain at least one of the two accompanying tugs until the anchorage had been passed, in case the steering problem should recur; (c) Failure to have the anchors ready to let go; (d) Failure to man the trick wheel while leaving harbor; and (e) Proceeding at a speed which was excessive in light of the prior loss of steering and the temporary nature of the repairs. The faults listed above are, of course, interrelated, and the faults of the harbor pilot are the faults of the master. See United States v. The WESTERBEL, 135 F.Supp. 596, 599 (S.D.N.Y.1955). On April 17, 1973, less than two months before the collision, the SEA WITCH had suffered a loss of steering. The repairs made to correct the problem were considered only temporary. PTO/UF 65; Exh. 107. If the master had informed the harbor pilot that there had been a recent loss of steering and that only temporary repairs had been made, it is doubtful that the harbor pilot would have taken the vessel to sea. Tr. 104. In any event, the harbor pilot testified that he would have kept the tugs alongside the SEA WITCH. Id. Failure to advise the harbor master constituted negligence. Moreover, the master, who was aware of the recent steering failure, should have kept at least one tug alongside the SEA WITCH until it was safely past the anchorage. Exh. 295, p. 1363; Exh. 303, p. 1548; Exh. 305, pp. 1925-26. The master, however, failed to do so. It is admitted that the port anchor could not be dropped. Prior to departure, the crew failed to raise the riding pawl, which was jammed against the anchor chain. Failure to ensure that the anchor was ready for release constituted a fault which contributed to the collision. See The SUNNYSIDE, 251 F. 271 (2d Cir.1918); River Terminals Corp. v. United States, 121 F.Supp. 98, 104 (E.D.La.1954); The WIRELESS No. 1, 290 F. 239 (E.D.N.Y.1923); Exh. 293, p. 1671; Exh. 320(b), pp. 2156-57; Tr. 107. The most serious precautionary fault, however, was the failure of the master to have the trick wheel in the steering engine room manned while the vessel was under way and until safely beyond the anchorage and harbor traffic. The trick wheel was unaffected by the steering failure and could have been used to steer the vessel. It is significant that on a prior occasion involving the steering mechanism of SEA WITCH, the same master had ordered that the trick wheel be manned when the vessel entered the harbor. Tr. 1809, 1817-18, 1839-40. The master’s failure to do so on the later occasion was negligent, as was his failure to direct that the trick wheel be manned once the loss of steering became apparent — as discussed hereinafter. The charge of excessive speed prior to the loss of steering is difficult to sustain in view of the good visibility and minimal traffic at the time. Absent the prior experience with the steering mechanism, the SEA WITCH’s speed would not have been excessive. However, in light of the prior steering problem — and the master’s knowledge of it — SEA WITCH’s speed must be considered excessive. Faced with the temporary nature of the repairs, SEA WITCH should have proceeded at a slower speed. Thus, under the circumstances, SEA WITCH’s speed was not reasonable. 2. Remedial Actions AEIL was also negligent in its navigation and seamanship of SEA WITCH after the loss of steering. The almost total failure to take remedial action to avoid or at least diminish the severity of the collision with ESSO BRUSSELS constituted negligence and was a proximate cause of the collision. The faults charged are: (a) Failure of SEA WITCH to stop her engines immediately upon loss of steering; (b) Failure of SEA WITCH to reduce speed or go full astern until the collision was inevitable; (c) Failure to order the trick wheel manned after loss of steering was known; (d) Failure to drop the starboard anchor. In order to understand the faults listed here and the extent of AEIL’s negligence, it is crucial to understand what took place during the time between the loss of steering and the collision. It is undisputed that steering was lost when SEA WITCH was changing course from 167° to 158° shortly after she passed Buoy 22. PTO/UF 10. The distance between Buoy 22 and ESSO BRUSSELS was approximately 1.4 miles. Professor Hamlin, a naval architect of vast experience, developed a track of SEA WITCH using a formula created by Professor Numoto. Exh. 328; Tr. 435. In developing this track, Professor Hamlin took into account the characteristics of SEA WITCH, the change of course from 167° to 158° during which the rudder froze in a position of 12° right, the speed of the vessel, the tide, and the prevailing currents. The track presented by Professor Hamlin was not made on a chart with reference to any geographical position but, rather, was a theoretical track in space, which Professor Hamlin placed over the chart upon which he had plotted the known position of the ESSO BRUSSELS. Tr. 461; Exh. 327. Because the collision occurred at 0042, Professor Hamlin moved the track so that the 0042 position was in contact with ESSO BRUSSELS and the upper position was on a heading of 167°. Tr. 434, 440-441. He found that this theoretical track, when so applied, showed that the change of course from 167° to 158° occurred shortly after passing Buoy 22 at a distance of approximately 1000 feet. Professor Hamlin found that the time between the loss of steering and the collision was 4 minutes and 23 seconds, and the distance covered was 6,290 feet, i.e., a little over a mile. Tr. 444-447. The validity of the formula employed by Professor Hamlin has not been seriously challenged. The Court does not find the testimony of the harbor pilot to be credible, nor does the Court accept the conclusions of AEIL’s expert witness. Professor Hamlin’s conclusions are consistent with the testimony of disinterested eyewitnesses who testified concerning the collision and the movements of SEA WITCH during the period immediately preceding the collision. As previously indicated, the Court has adopted the findings of Professor Hamlin with respect to the track of SEA WITCH during the period after passing Buoy 22 and ending with the collision with ESSO BRUSSELS. Based on Professor Hamlin’s testimony, together with the other evidence presented in this case, the Court is convinced that, after the loss of steering, SEA WITCH had sufficient time to prevent the collision by taking remedial actions. The failure to take such actions constituted negligence. See Complaint of Flota Mercante Grancolombiana, S.A., 440 F.Supp. 704, 714 (S.D.N.Y.1977); see also The NEW YORK, 175 U.S. 187, 209, 20 S.Ct. 67, 75, 44 L.Ed. 126 (1899); The BERN, 74 F.2d 235, 237 (2d Cir.1934); The COMMERCIAL MARINER, 63 F.2d 798, 800-01 (2d Cir.1933), cert. denied, 290 U.S. 643 (1933); The CHEROKEE, 45 F.2d 150, 151 (2d Cir.1930). The failure to stop engines constituted the most serious navigational fault that occurred after steering was lost. Although the harbor pilot testified that he did stop the engines almost immediately, the Court does not find such testimony credible in light of his Coast Guard testimony. His testimony also lacks credibility in light of the testimony of Penswick, mate of the tug BARBARA MORAN, and the computations of Professor Hamlin. It is clear that the engines of SEA WITCH were not stopped when steering was lost, but remained at full ahead until collision was a certainty. This was a serious fault. It has been suggested that the harbor pilot was misled when he mistakenly assumed that the rudder order indicator— which was in the full left position — was the rudder angle indicator, i.e., that the rudder itself was full left instead of being in a locked position of 12° right. Tr. 78-81. However, once the helmsman had reported that the ship was not steering, there was no possible excuse for not stopping the engines. Proceeding at full ahead was gross negligence. It was also gross negligence not to man the trick wheel immediately upon learning that the steering was suspect. As previously indicated, the captain had more than four minutes to act between the loss of steering and the collision. That was more than enough time to man the trick wheel and control the direction of the SEA WITCH. Failure to do so clearly violated the principles of proper seamanship under the circumstances of this case. Finally, the starboard anchor should have been released. Dropping that anchor might have prevented the collision by slowing, stopping, or turning the vessel. Even though the port anchor was fouled, there was no excuse for failing to drop the starboard anchor. Failure to release the starboard anchor until the moment of collision was at least a factor that contributed to the casualty and it constituted negligence on the part of AEIL. B. APPORTIONING LIABILITY In United States v. Reliable Transfer Co., Inc., 421 U.S. 397, 95 S.Ct. 1708, 44 L.Ed.2d 251 (1975), the Supreme Court abolished the doctrine of divided damages in maritime collision cases and adopted the principle of proportionate liability. The Supreme Court held that: when two or more parties have contributed by their fault to cause property damage in maritime collision ..., liability for such damage is to be allocated among the parties proportionately to the comparative degree of their fault, and ... liability for such damages is to be allocated equally only when the parties are equally at fault or when it is not possible fairly to measure the comparative degree of their fault. Id. at 411, 95 S.Ct. at 1715-16. In the instant case, the Court has made sufficient findings, many of which are undisputed, to measure the comparative degree of fault as between AEIL and the defendants involved herein. As discussed above, the ESSO BRUSSELS, which was anchored at the time of the collision, bears no responsibility for the casualty. AEIL, however, must bear at least partial responsibility for the casualty, in light of its negligent seamanship and navigation, which was a proximate cause of the collision. Because the loss of steering was also a proximate cause of the collision, liability for the damage sustained must be allocated proportionately based on the comparative degree of fault that exists. The Court concludes that 60% of the responsibility for the collision must be attributed to AEIL’s navigation and seamanship. If AEIL had taken appropriate remedial and precautionary measures, there would have been no collision. See Pan-Alaska Fisheries, Inc. v. Marine Const. & Design Co., 565 F.2d 1129, 1139 (9th Cir.1977) (“[T]he award of damages shall be reduced in proportion to the plaintiffs contribution to his own loss or injury.”); see also Flota Mercante Grancolombiana, 440 F.Supp. at 726 (The court found that proper navigation and seamanship would have prevented the collision, and therefore allocated 82.5% of the responsibility for the collision to the ship that lost its steering). It is not the number of faults that matters, but rather it is the nature of the faults that is conclusive. In his computations, Professor Hamlin described the course SEA WITCH would have followed, if the captain had executed certain orders: (1) If a stop order had been made 30 seconds after loss of steering was known, the SEA WITCH would have missed the ESSO BRUSSELS and stopped several hundred feet to the east of her. Exh. 332; E on Exh. 327. (2) If a stop order had been made one minute after loss of steering was known, the SEA WITCH would have missed the ESSO BRUSSELS and stopped southwest of the ESSO BRUSSELS. Exh. 333; E on Exh. 327. (3) If emergency full astern had been ordered one minute after loss of steering was known, the SEA WITCH would have stopped before ever reaching the ESSO BRUSSELS. Exh. 331; C on Exh. 327. (4) If emergency or normal full astern and drop the starboard anchor had been ordered one minute after loss of steering was known, the SEA WITCH would have stopped well before reaching the ESSO BRUSSELS. Exhs. 329, 330; A and B on Exh. 327. The plaintiff, AEIL, argues that the Court should not “second guess” the captain and substitute its judgment, based on hindsight, for the captain’s judgment. In this case, however, the Court has not substituted its judgment for the captain’s. The captain failed to exercise his judgment; he failed to take any action to prevent the collision. The Court has merely concluded, as it must, that the captain’s failure to exercise his judgment and to take remedial action constituted negligence. Although AEIL should have acted to prevent the collision, there would have been no collision if the steering had not failed. Indeed, there would have been no need for AEIL to take remedial measures to avoid the collision if the steering had not failed. Thus, AEIL cannot be held solely responsible for the collision. For the reasons set forth above, the Court has apportioned responsibility for the collision as follows: AEIL’s negligent navigation and seamanship — 60%; loss of steering — 40%. PRODUCTS LIABILITY The Court has determined that the navigational error on the night in question was not the sole cause of the collision, and that the failure of the steering system represented 40% of the cause of the collision. Therefore, the Court will proceed to make factual findings and conclusions of law regarding AEIL’s products liability claim against Bath. In essence, AEIL asserts that Bath should be found liable for defective manufacture and design of the steering system, and for failing to provide adequate warnings. FINDINGS OF FACT A. THE CONSTRUCTION OF THE SEA WITCH 35. During 1964, AEIL decided to have four cargo ships built. AEIL retained John J. McMullen Associates, Inc. (“J.J. McMul-len”), a naval architecture firm, to prepare the bidding specifications and plans showing the general nature and characteristics specified by AEIL for these vessels. AEIL ultimately decided to build only three of these ships. PTO/UF 22. 36. These bidding specifications and plans were presented to the United States Maritime Administration (“MARAD”) for approval. PTO/UF 23; Exh. 30. 37. MARAD approved the bidding specifications, and they were sent to several shipyards which were invited to submit bids. PTO/UF 24. 38. The bidding specifications were prepared by J.J. McMullen based on consultation with and general guidance from AEIL. Tr. at 296 (Knight). 39. A description of the steering system was contained in the specifications, but the specifications did not include a description of the linkage between the Sperry rotary power unit and the differential gear box. Exh. 30, at 81-9, 94-5 to 94-7; Tr. at 285-286, 294. 40. The bids submitted were forwarded to MARAD for review and approval. PTO/UF 25. 41. The contract for the construction of the SEA WITCH was awarded to Bath. It was executed on November 30, 1965 by MARAD, AEIL, and Bath. PTO/UF 25, 26. 42. Hyde Windlass Company (“Hyde”), which at that time was a division of Bath, furnished the unitized hydraulic rudder control system which was installed in the steering gear room. This steering system included the linkage between the differential gearbox, as well as the RPUs — the rotary power units — which were made by Sperry. PTO/UF 29. Sperry did not design, manufacture, or install any part of the linkage connecting the RPUs to each other or to the differential. PTO/UF 57. 43. The linkage system connecting the Sperry RPUs to the differential gearbox was tested more than a dozen times in the Hyde shop. Tr. 1614 (Witham). 44. On or about July 25, 1968, dock trials of the SEA WITCH were conducted by Bath. During the dock trials, representatives of various concerns, including AEIL, MARAD, the Coast Guard, and Bath, tested various equipment, including the steering gear. The test of the steering gear was satisfactory. PTO/UF 41. 45. On or about August 5, 1968, builder’s trials were conducted of the SEA WITCH at sea. During the builder’s trials, representatives of various concerns, including AEIL, MARAD, the Coast Guard, and Bath were present. Various equipment was tested, including the steering gear. The test of the steering gear was satisfactory. PTO/UF 42. 46. On August 14 and 15, 1968, the SEA WITCH underwent her Official Sea Trials, in accordance with a trial agenda previously approved by AEIL and MAR-AD. Representatives of AEIL, Bath, Hyde, Sperry, MARAD, and the Coast Guard, among others, were present aboard the SEA WITCH during the Sea Trials. The test of the steering gear was satisfactory. PTO/UF 43. 47. On September 4, 1968, after further testing of the steering system in port under the supervision of AEIL and MARAD representatives, delivery of the SEA WITCH was accepted by AEIL. PTO/UF 44; Exhs. 75, 76. Only five exceptions were noted by AEIL, none of which involved the steering system. Tr. 995 (Mus-senden); Exh. 77. 48. The guarantee under the Construction Contract ran from September 4, 1968 through March 4, 1969. Exhs. 93, 94. On March 13, 1969, the final guarantee survey was completed. Exh. 88. No objections were registered regarding the steering linkage or other steering system components. Id. 49. On August 3, 1973, MARAD, AEIL, and Bath executed a release which stated that, upon payment of the amounts indicated therein, the Construction Contract would be “completely and finally settled and all obligations of the Parties thereto shall be considered accomplished.” Exh. 153; see also Exhs. 116, 117, 122. 50. At no time prior to the casualty of June 1-2, 1973 was Bath either consulted or informed about any alleged problems with the steering system of the SEA WITCH. Tr. 6 (indicating that this fact is conceded by the parties). B. THE LINKAGE BETWEEN THE DIFFERENTIAL AND THE RPUs 51. The port RPU was connected to the differential by a linkage consisting of a connecting shaft between two universal joints (or “universals”). PTO/UF 56; see Exh. 187(c). 52. This linkage had been installed by Hyde in twelve other vessels prior to the construction of the SEA WITCH. Tr. 1494-1495 (Sturtevant). 53. The linkage is depicted in Plan No. 20561, prepared by Hyde and approved by AEIL on January 12, 1967. Exh. 50-51; Tr. 1352-53 (Bowles); see Exh. 200; see also Exh. 219, at Fig. 1-8. 54. Plan No. 20561 portrays, inter alia, the three keys and keyways called for on the forward universal and the forward hub of the aft universal. All three keys are straight and uncaptured. Exh. 200; see Exh. 219, at Fig. 1-8. 55. The forward hub of the universal closest to the RPUs was affixed to the port RPU output shaft by a square key 2V8* long and 8/i6" wide and a set screw 90° from the keyway. Exh. 200; see Exh. 219, at Fig. 1-8. 56. The aft hub of this universal joint was affixed to the connecting shaft by a feathered square key n/i6" long and 3/i6" wide, which was pinned to the connecting shaft by a roll pin Vs" long. No set screw was installed at this point. This type of connection point is generally called a slip joint, or expansion joint. It is designed to permit the linkage to “breathe,” and to relieve axial motion, and alleviate stresses caused by expansion, contraction, and vibration. See Tr. 1607-1608 (Witham); Tr. 753-755 (Bond); Tr. 1081-1085 (Freuden-stein). 57. The connecting shaft was fastened to the input shaft of the differential gear box by another universal joint (“the aft universal”). The forward hub of the aft universal was affixed to the connecting shaft by a square key lVi6" long and 3/i6" wide and a set screw 90° from the keyway. Exh. 200. 58. The aft hub of the aft universal was inserted over the input shaft, or stub shaft, of the differential. PTO/UF 58. 59. As originally constructed by Hyde, this hub was secured to the stub shaft by a Number 3 Woodruff key, which, according to machinists’ standards, is a half round piece of metal Vs inch thick and V2 inch in diameter. PTO/UF 59. This Woodruff key can be visualized as being somewhat similar to a half-moon or crescent shape. 60. A semicircular keyway, approximately Vs inch wide, lk inch in length, and Vs inch in depth at its deepest point, was milled longitudinally along the stub shaft in its center. 61. A key way which was approximately Vs inch wide was milled from end to end in the aft hub of the aft universal. PTO/UF 62. 62. A set screw was installed in the aft hub of the aft universal joint approximately 90° clockwise (looking forward) from the keyway perpendicular to and exerting pressure upon the stub shaft. PTO/UF 62. 63. When the Woodruff key was inserted into the stub shaft (as if half a dime were inserted, curved side down and lengthwise into that stub shaft) and the universal hub keyway was placed over the portion of the Woodruff key that was not submerged in the stub shaft keyway, the semicircular keyway in the stub shaft prevented the Woodruff key from sliding in a longitudinal direction out of the stub shaft, and the universal hub prevented it from moving up or down perpendicularly in relation to the stub shaft. PTO/UF 61. C. INSPECTION AFTER THE COLLISION 64. As soon as there was access to the steering gear room, the SEA WITCH steering system was inspected by the Coast Guard and others. That inspection showed that there was no damage to the linkage assembly between the RPUs and the differential because of the collision. The RPUs were found to be at hard left rudder in accord with the last steering command which had been given from the bridge. This order did not, however, reach the rudder, which was found to be in a position of 12° right. PTO/UF 18, 19; Tr. 1132. 65. A 3/i6 inch straight square key which had been in the differential stub shaft was no longer in the stub shaft keyway, but was lying in the bottom of the connection between the two hubs of the aft universal. PTO/UF 20. This straight key was not part of the original steering gear as delivered to AEIL. It had been installed by AEIL on April 23, 1973, when the stub shaft was modified by eliminating the Woodruff key, and replacing it with the straight key as further described below. 66. The absence of this straight key from its position in the keyway of the aft universal hub and the differential' stub shaft permitted the shaft connecting the RPUs to the differential to rotate without transmitting that rotation to the stub shaft of the differential. This caused the steering failure on June 2, 1973. PTO/UF 21. D. PRIOR HISTORY OF THE SEA WITCH 67. From the date of delivery of the SEA WITCH to AEIL in 1968 until the date of collision in 1973, the SEA WITCH was continuously operated by AEIL in its transatlantic liner service with the exception of (1) the time spent on routine shipyard maintenance and repairs, and (2) the period of October 17, 1970 through January 13,1971, when the vessel was out of service for extensive repairs and modification to her hull in Brooklyn, New York. PTO/UF 63. 68. Between the date of delivery and the date of collision, the ship had lost steering control on two occasions because of a failure of the linkage between the Sperry RPUs and the differential. Both failures occurred on July 8, 1971. PTO/UF 64. 69. The steering loss in July 1971 occurred when the square key which belonged in the forward hub of the aft universal fell out. Approximately four hours after the ship’s Chief Engineer had reinserted the key and tightened the set screws that had loosened, the same key again fell out of the linkage and steering was lost for a second time. The Chief Engineer then added additional locking set screws over the original set screws. That part of the linkage connection experienced no further failures. Tr. 1809-1815. 70. AEIL failed to show that the linkage components in place in July 1971 were delivered with the system. In addition, AEIL failed to show that such a failure was possible with linkage components of the original size. Thus, the Court cannot find that this failure resulted from, or evidenced, a defect in the steering system. i) The Woodruff Key and Keyway on April 17, 197S 71. On April 17, 1973, the SEA WITCH experienced a steering problem that result•ed from a leaking relief valve in the steering gear hydraulic system. After performing a temporary repair on the relief valve, which corrected the problem, Chief Engineer Thomas Hind (“Hind”) and First Assistant Engineer Clifford Smith (“Smith”) checked over other parts of the steering gear and discovered slight play between the stub shaft of the differential and the aft hub of the aft universal. PTO/UF 65. 72. When the connection was disassembled on April 17, 1973, it was discovered that the Woodruff key and the stub shaft keyway were worn. PTO/UF 66. 73. The key was almost rainbow shaped, arched, and all of its knife edges were gone. Tr. 574-575 (Smith). The Woodruff keyway was also damaged — it “had no knife edges and no sharp corners .... [0]n the long side it was football shaped. On the ends it just wasn’t square. It wasn’t square in any direction looking at it directly from the top_” Tr. 175 (Smith). 74. Smith obtained an undamaged Woodruff key, possibly from another piece of machinery aboard the ship, and this substitute was inserted into the keyway. Tr. 600; PTO/UF 66. It appears that the worn key was then thrown away. Tr. 599-600. 75. Smith also added a set screw hole to one of the universal joints. Tr. 612-613. This additional set screw was not provided for in the Hyde design. See Exh. 200. 76. Although there was no conclusive evidence presented regarding the cause of the wallowing of the keyway observed on April 17, 1973, the Court draws the inference proposed by Bath’s expert Professor John K. Tien (“Prof. Tien”) that the wallowing was caused by the use of a set screw in the Woodruff keyway without a key. Prof. Tien concluded that at some point prior to April 17, 1973, AEIL had inserted a set screw through the aft hub of the aft universal into Hyde’s original Woodruff keyway. The insertion of a set screw deformed the keyway. The operation of the steering linkage using that set screw in the keyway instead of a key further wallowed out the Hyde keyway into the “football shape” which Smith observed when he took apart the linkage and examined the keyway in April 1973. Tr. 1663-1668, 1674-1675, 1678-1679, 1684, 1694 (Tien); see also Tr. 1457 (Sturtevant). Plaintiff’s expert, Dr. Vittorio Castelli, allowed that deformation of a key could occur as a result of the key being placed in a wallowed keyway. Tr. 1979-1980. The Court finds that the evidence supports the inference that the aft universal hub was at some time prior to April 17, 1973 linked to the differential stub shaft by means of a set screw run into the keyway itself. The evidence tended to show: (1) that the Sperry stubshaft was operated in this manner, i.e., by using a set screw in the key way rather than a key, see Exhs. 363(a)-363(d); Tr. 1676 (Tien); see also Tr. 1077-1078 (Freudenstein); (2) that a set screw had been inserted through the original hole drilled by Hyde in the aft universal hub and into a new hole drilled by AEIL in the differential stub shaft, and that no Woodruff key was in the key way during this arrangement, see Tr. 1669-1673 (Tien); Exh. 362(c); (3) that a set screw had been inserted into the aft universal hub and onto the stub shaft in a location which made it physically impossible for a Woodruff key to be in the keyway at the same time. Tr. 1673-1674 (Tien); Exhs. 362(f), 362(g). 77. The Court also draws the inference that the linkage had been disassembled by AEIL on several occasions. This is evidenced by the various locations of set screw marks made on the linkage’s shafts. See Tr. 1668 (Tien); Tr. 1079 (Freuden-stein); see Exhs. 362(b), 362(c), 362(f). 78. Defendant has collected substantial circumstantial evidence, persuasive to this Court, that, in addition to replacing the Woodruff key with a straight key, see infra Till 85, 89, 108, 112, 116, and adding a set screw hole to one of the universal joints, see supra ¶ 75, AEIL made other changes in the steering linkage which departed from the original design. These changes include: (1) elimination of the Hyde linkage’s feather key, see Tr. 1607-1608, 1627 (Witham); Tr. 1454-1456, 1531 (Sturtevant); Exhs. 187(c), 196, 196(a); (2) replacement of the linkage’s long connecting shaft, see Tr. 1607-1608 (Witham); Tr. 1455-1456, 1531 (Sturtevant); Tr. 1879 (Jo-hansen); Exhs. 204,187(c); (3) addition of a set screw to the aft universal, see Tr. 1608-1609; Exhs. 187(b); (4) addition of a set screw to the forward universal, see Tr. 1556 (Sturtevant); Exhs. 187(d), 196, 196(a); and (5) countersinking of set screw holes in the connection shaft’s straight keys, Tr. 1830-1832, 1835, 1839 (Johansen); Exh. 187(f). 79. In any event, the replacement key did not fit perfectly in the keyway, and this repair was regarded as only a temporary measure. PTO/UF 67. 80. Hind and Smith had discussed the type of repair that should be made when the ship returned to New York. As described by Smith, “the discussion was about whether there was a better key to put in there than a Woodruff, meaning one with more strength, and it was pretty well agreed upon between he and I, at least, that a straight key would give you a lot more strength.” Tr. 577-578. 81. Neither Hind nor Smith was qualified to make design changes to the SEA WITCH’s steering linkage. Tr. 646-647, 650 (Hurd); Tr. 602 (Smith). 82. On April 19, 1973, the SEA WITCH, while still at sea, radioed AEIL’s New York office with a list of repairs. PTO/UF 68; Exh. 136. 83. When the radio message was received by AEIL, the responsibility for seeing that the repairs were completed was assigned to an AEIL Port Engineer, William Hurd (“Hurd”), PTO/UF 69, who was told that he would have to arrange for these repairs when the vessel arrived in New York around April 23. Tr. 631-632 (Hurd). 84. Hurd was not familiar with the steering system of the SEA WITCH, and had never gone into the steering gear room of any SEA WITCH class vessel prior to April 1973. Tr. 630. Hurd was not a designer of marine equipment or expert in-mechanical linkages, and did not know what considerations went into the design of the SEA WITCH steering linkage’s various components. Tr. 646-647, 650 (Hurd). ii) Repairs by Bond Hydraulic on April 23, 1973 85. On or about April 20, 1973, Hurd telephoned Eugene Bond (“Bond”), of Bond Hydraulic, and informed him that the SEA WITCH had a Woodruff key that was worn and that a repair had to be effected. PTO/UF 70. 86. Bond is the owner of Bond Hydraulic, which is in the business of repairing marine hydraulic equipment. From approximately 1950 to the mid-1960s, Bond worked with the Hyde Windlass Company. Tr. 701-702. Bond was not a designer by training or experience. Tr. 737. 87. In the conversations between Hurd and Bond, Hurd stated that Hind would like to have a square key put into the shaft, and asked if it was feasible. No definite decision to insert a straight key was made prior to the arrival of the SEA WITCH in port. Tr. 790, 784, 787. 88. Neither Hurd nor Bond was certain of the problem on the SEA WITCH at the time of the first communication between the two. Bond thought that the gear box involved could be either of two possibilities. Bond told Hurd that he needed more details. Hurd said he would contact the ship to see if he could obtain further information. Tr. 718-719 (Bond). 89. Later the same day Hurd called back and informed Bond that he had gotten a wire, and that the part needing repair was the stub shaft, or input shaft, on the differential gear box. Tr. 722 (Bond). 90. Meanwhile, Bond had called Claude Clark (“Clark”) who worked for Hyde-Zim-mie Products — an Ohio company which had bought out Hyde — and asked how long it would take to deliver a differential input shaft and the airborne miter gear box. Tr. 719-720. Bond hoped that Clark could send the materials to him by air so that he would get them before SEA WITCH arrived in port. Tr. 721. Clark said that he did not have either piece in stock, and that there would be “roughly two weeks’ delivery on this material.” Tr. 720-721. Bond said he could not wait that long. Bond testified that “[t]hat ship was coming in Monday morning and as [f]ar as [he] knew, if it was on a normal schedule it would be leaving Monday night, and [he] didn’t have time to wait two weeks.” Tr. 721; see Tr. 868, 882. 91. Bond next tried to locate a milling machine cutter that would cut the groove for the No. 3 Woodruff key. Tr. 722. Bond was not successful, nor was his son, Bobby Bond (“Bobby”), who tried further the next day, contacting a number of establishments. Tr. 722, 812-813. There is conflicting evidence as to how long it was expected to take to locate such a machine— whether a couple of hours or days. Tr. 805-815. But it appears to the Court that the time which Bond estimated would be required — when speaking to Hurd about the type of repair which should be undertaken — was no more than four or five hours. Tr. 814, 806. As it turned out, a cutter was located a couple of days after the repair. Tr. 723, 846-847. 92. Bond recommended to Hurd that the shaft be replaced with another shaft fitted with a Woodruff key. Tr. 737 (Bond). 93. Bond would have preferred to retain the original feature, and knew that the proper way to handle the repair would have been to put in a new shaft with a Woodruff key. Tr. 795-796. Indeed, Bond testified that his “policy has always been to use genuine parts.” Tr. 795. But, he noted, “when parts are not available and the ship is leaving, you have to do something.” Id.; Tr. 797. Bond also stated that in making such repairs, “[i]t’s better to keep the design and the intent of the original design engineer unless they themselves change that design.” Tr. 798. 94. Bond stated, in his testimony before the Coast Guard in 1973, that the repair was a case of necessity: since the vessel was sailing in about six hours, it was necessary to change the design, as opposed to pursuing other options. Tr. 804. 95. The Court finds that it is more likely than not that AEIL rejected Bond’s recommendation to retain Hyde’s captured Woodruff key design primarily because AEIL did not want to risk delaying the SEA WITCH’S departure. Tr. 804, 805-806 (Bond). 96. Hurd provided Bond with a Xeroxed section of Fig. 1-7 of the Hyde manual, which depicted the area in which the work was to be done. Tr. 788. Hurd did not send Bond Fig. 1-8, and, although Bond had a copy of the Hyde manual in his shop, it appears that he did not consult this drawing before the repair. Tr. 788-789. Bond testified that Bobby did not use Fig. 1-8 in manufacturing the key. Tr. 829. 97. Bond’s testimony was to the effect that he had not referred to any manual or drawing with respect to the alteration of the Woodruff key to a straight key. Indeed, Bond testified that Fig. 1-8 does not indicate that there should be a straight key in the location of the Woodruff key. Tr. 789. 98. Bond knew that the original Hyde design and installation called for the use of a Woodruff key. Tr. 789-790 (Bond). 99. Although Hurd testified that he considered Fig. 1-8 in the Hyde manual, and interpreted it to mean that a straight key was contemplated in the aft hub of the aft universal, Tr. 657, Hurd also testified that he knew that a Woodruff key must actually have been installed in the ship originally. Tr. 667 (Hurd); Tr. 600-602 (Smith). Thus, the Court finds that Hurd was not, as a matter of fact, misled by Fig. 1-8 in the Hyde manual. Nor was evidence produced to suggest that anyone else was, as a matter of fact, misled by the drawing. Thus, the Court holds that Fig. 1-8 of the Hyde manual did not cause any misunderstanding regarding the original Hyde design. 100. Although Bond states that he rejected the possibility of replacing the original .shaft primarily because he did not know what kind of material the shaft was made out of, the accuracy of this statement is called into question by the fact that Bond himself concedes that he could have selected a steel whose strength was certain to be adequate, and safely used it for a new shaft. Tr. 848; see also Tr. 724-725. Therefore, the Court cannot give weight to Bond’s statement that his lack of information about the steel influenced his view of the options available with respect to the repair. 101. Bond stated that, because the %6 inch square keys in the Sperry output shaft and in the control shaft seemed to be holding up better than the Woodruff key, the best way to handle the repair would be to put a %6 inch square key in the input shaft of the differential as well. Tr. 727. 102. On April 23, 1973, the SEA WITCH arrived in port in New York. PTO/UF 72. 103. A requisition for material was prepared by Hind and was transmitted to AEIL upon the SEA WITCH’s arrival in New York. PTO/UF 71; Exh. 135. 104. After the vessel’s arrival, Hurd went aboard and spoke to Hind and Smith who told him that there was a Woodruff key in the aft universal hub that was worn excessively, and that they would like to have it replaced with a straight key. Tr. 657. 105. When the vessel arrived, Bond was in Rhode Island, and Bobby went aboard. Tr. 724. 106. Bond had a conversation with Hurd when Bobby was still on the ship, but Hurd did not give Bobby any instructions based on the phone conversation. Tr. 820, 822. 107. It appears that an initial determination that a straight key would be inserted was made before Bobby left the ship. Tr. 794; 658-659. However, in Bond’s conversation with Hurd after the parts had been brought to the shop, the decision was made definite. Tr. 791. 108. The stub shaft along with the gear affixed to it, and the aft hub of the aft universal were removed and brought to the Bond Hydraulic workshop. PTO/UF 72. 109. As observed by Bond at his shop, the Woodruff keyway had wallowed out at the sides. The keyway in the shaft was worn in the area of the key — by approximately ¥i6th of an inch on either side of the original edges of the keyway. The keyway in the universal was worn somewhat, but not as badly as that in the shaft. Tr. 724, 726-727. 110. Bond spoke to Hurd on the day the parts were taken off the ship. After this conversation with Hurd, the work was done on the parts. Tr. 728. 111. Hurd did not ask for Bond’s endorsement of the proposal to replace the Woodruff key with a straight key. Tr. 867. 112. The unit was disassembled and a square keyway approximately %6 inch wide and %2 inch deep was milled to the end of the existing stub shaft. The keyway in the universal hub was widened to approximately 3/ie of an inch. PTO/UF 74. 113. Bobby Bond did the work involved, i.e., milled out the parts and assembled the pieces. Tr. 728. 114. Bobby Bond was then twenty-two years old and had been working for Bond Hydraulic for three years. He was not an expert in either engineering or metallurgy. Bobby was not a design engineer of any sort nor did he have machinist’s credentials. He was a machine operator who, according to Bond, was “good on the machine.” Tr. 817, 817-818. 115. Bond did not supervise the repair done by Bobby Bond, nor did he give him instructions as to how to do the job. Tr. 822-23 (Bond). 116. Bond Hydraulic also made a key which was s/i6" square and 2%2" long. This key was inserted in the stub shaft keyway. PTO/UF 75. 117. Bond did not examine the key after it was made. Tr. 824. 118. Bond testified that he did not know how much pressure was required to insert the key Bobby had made into the keyway. Tr. 856; see also Tr. 857. He also testified that a substance called “Lock-tite” was placed on the universal coupling before it was assembled. Tr. 780-783; 826-827. Bond conceded, however, as Bath’s expert testified, that use of this substance was not the proper way to secure a key in an open keyway. Tr. 827; Tr. 1725-1726 (Tien). 119. After reassembly, the linkage was reinstalled aboard the SEA WITCH. PTO/UF 75. 120. Bond testified that only qualified design engineers have the qualifications to change someone else’s design. Tr. 798. 121. “[T]he error in this fixture,” according to Bond, “is ... [that] it was not a captive key. Had that been a captive key, we would not be sitting here today, and [the repair with the straight key] would hold up better than the Woodruff key that was in there [before].” Tr. 811. Bond testified further “[w]hat [he] ... should have done here ... is to cut a double-ended key with a key seat so that it was captured.” Tr. 816. 122. A Woodruff key, or even some other kind of captured key — such as a sunken key or a pinned straight key — would not have fallen out as did the uneaptured straight key inserted on April 23rd. Tr. 799-800. A sunken key is one which is inserted into a keyway milled out just short of the shaft’s end so that there is metal at both ends of the keyway keeping the key from coming out. Id. 123. A pinned straight key involves drilling a small hole through the key and putting a pin through it, thus fastening the key firmly to the stub shaft. It would have been preferable to use a captured key in the repair, and it would have been no more difficult or time consuming than the repair actually performed. Tr. 799-801, 827 (Bond); Tr. 1472 (Sturtevant); Tr. 681-682 (Hurd). 124. The Court finds that the square key installed by Bobby Bond was loose in the keyway at installation, and was not, as AEIL contends, tight-fitting. Tr. 764-765 (Bond); Tr. 1096, 1118 (Freudenstein); Tr. 1651-1655,1662-1663,1658,1716 (Tien). At trial it was demonstrated that the key now fits loosely into the keyway, and requires no pressure for its insertion into the key-way. Tr. 764-765 (Bond); Tr. 2062 (Crank-shaw). Further, Bond’s testimony to the contrary must be rejected, since it appears that he did n