September 20, 1988
Mr. Ronald L. Schleede Chief, Aviation Accident Division National Transportation Safety Board Washington, D.C. 20594
Dear Mr. Schleede
Thank you for your letter dated August 3, 1988. I have carefully studied the Aloha Hearings Transcript and Exhibits (Seattle, July 12 to 15, 1988) and, with them in hand, have re-studied the Far Eastern accident report and Appendices. I have again concluded that the Far Eastern aircraft, like the Aloha aircraft, rapidly decompressed at the upper level of the forward fuselage. In fact, it now appears that the accidents are more similar than previously thought.
My comments at Filings #1 dated June 27, 1988 focus on the Far Eastern breakup patterns indicated by the drawing (of the forward fuselage skin breakup) At Chinese Appendix 19, which is omitted from the Appendices in the English Report. My comments here focus on breakup details appearing in the Appendices to the Report, but which the report itself disregards, which relate mainly to the separation sequence, the floor beams, the floor beam upper flanges, and their respective fracture and damage patterns. In fact, we see that the Report with the Appendices, and the Report without the appendices, give quite different pictures of the decompression and ensuing breakup. We also see that the prevailing view (that the decompression occurred at the lower lobe bilge area) is erroneous, provided we carefully read and study the Appendices. I have now done just that.
I submit the following additional comments for your consideration.
Attachments hereto are (I) Contents of the Reports on the Far Eastern Accident, (II) Brief Chronology of the Investigating and Reporting of the Far Eastern Accident, (III) Descriptions of the Far Eastern and Aloha Floor Beams and Adjacent Structure, and (IV) Picture of the Far Eastern Separation Sequence.
Far Eastern Investigation, Reports, and Appendices
Aloha Exhibit 15D. This is the undated Accident Report in English ("English Report*) cited at Filings #1. Appendices 1 to 21 and, of course, Chinese appendix 19 are omitted. I personally mailed copies to the Airline Pilots Association at Herndon, Virginia, so the Appendices are now available in the Washington area in case your docket files are incomplete. Attachment I lists the appendices by subject and source. Without the Appendices, the Report is ambiguous and misleading in several important respects.
Aloha Transcript p. 492. In reply to Chairman Lauber, Mr. Harradine states that the English Report is a translation of the Chinese. However, any experienced translator can tell that the English Report is not a translation, but that the Taiwanese had the English in hand when preparing their Report in Chinese, Attachment II also indicates that Boeing and NTSB investigative activity continuing well into late March 1982 hardly permitted preparation of an original report in Chinese. Of course the Chinese is the official Report in Taiwan even though it is a modified translation of the English. (A true partial English translation of the Chinese report (parts 2 and 3 of the report only) was made in 1983. see Attachment I.)
I therefore know that Aloha Exhibit 15D is not a translation as purported, but I do not know who prepared it. However, in an Opinion dated September 28, 1987, Bernard C. Doyle states that Boeing wrote the English report for NTSB transmittal to the Taiwanese. I separately enclose a copy of the Opinion, At the time, Mr. Doyle, like the rest of us, did not question the prevailing view that the decompression occurred at the bilge area. Therefore, the Opinion does not touch on the SB 1039 skin lap condition now believed to be a third contributing cause
(possibly the initiating cause) of the breakup.
Aloha Transcript p. 491. In reply to Chairman Lauber, Mr. Purvis states that no one at Boeing has first-hand knowledge of the accident or investigation. However, the Doyle Opinion and attachments I and II indicate substantial Boeing involvement, much of it by Mr. Purvis himself.
The English Report (p. 35) attributes the breakup to corrosion and fatigue at the lower fuselage (lower lobe), but stops short of explicitly pin-pointing any particular location as the place of decompression. This is ambiguous and misleading, especially when so much space is devoted corrosion at the lower fuselage bilge area. With the Appendices, or without them, readers of the Report will focus differently and reach a number of different conclusions. Take, for example (see also below), decompression resulting from an upward skin tear starting from S-19L or BS 360 below S-17 floor level (at the lower lobe). and reaching critical length above S-17 floor level (at the upper lobe). Would the structural failure be a lower or upper lobe failure?
Chinese appendix 19 indicates upper lobe decompression. The separation sequence analyzed here also indicates upper lobe decompression, as do the floor beam breaking patterns and the damage to the upper flanges also analyzed here. The Appendices indicate that only 2 out of 11 floor beams have potential to indicate lower lobe decompression. However, because Exhibit 15D omits the appendices, these 2 out of 11 carried the day Seattle. The prevailing view is still with us. I trust that you will shortly agree that that view is simply erroneous.
Far Eastern Separation Sequence
English Report p. 2, 7, 11, and 22. The Far Eastern aircraft was flying in a north-south direction when the decompression occurred at 22,000 feet altitude. After the decompression, structural parts progressively separated and fell along the north-south flight path for 15 kilometers until the main fuselage hit ground. The north-south flight path is, of course, also the north-south
separation path of the aircraft. I cannot explain why, but the English Report (p. 7) ambiguously and misleadingly describes the separated structure (and separation path and sequence) in reverse south-north. The report does not admit to such reversal, but it describes (A) first the main fuselage at "location A", (B) then the nose section at "location B", and (C) last the many upper fuselage parts at "location C". (The Chinese report uses place names and is not misleading.)
However, appendix 4 indicates that (A) "location A" is at the end of the flight path where the main fuselage hit ground at km.15, (B) "location B" is at km.10 where the nose section fell after separating at about 9,000 ft., and (C) "location C" is the 10-kilometer stretch from km.1 where the aircraft decompressed, to km. 10 where the nose section separated, and where the forward fuselage parts progressively separated and fell along the flight path as the aircraft descended from 22,000 ft. to 9,000 ft. The Appendix also indicates that the Report fails to describe (.) parts that separated and fell at the 5-kilometer stretch from km. 10 to 15.
Appendix 11 (par. B.5) indicates that the nose section, which fell at km. 10 after separating at about 9,000 ft., separated by tipping downward, shearing off at BS 360 bottom, and carrying with it the electrical equipment compartment at BS 360 to 380 under s-17 floor level. This indicates that the nose section separated before the belly structure did. In turn, this indicates that the belly structure fell at the undescribed stretch from km. 10 to 15 after separating below 9,000 ft.
Reverse South-North Order Omitting the Belly Structure
The English Report (Report only) description in reverse south-north order, with other stretch from km. 10 to 15 omitted, and without details as to location or altitude, gives the following sequence.
(A) The aircraft "midsection of fuselage, empennage, wings, engines and main landing gears ... fallen in one piece" ............. (.)....................................................................................................
(B) The "nose section that included the entire cockpit, electronics equipment compartment, nose gear and both left and right cabin doors ........struck the ground" .........................................
(C) The "forward section of the fuselage ...... forward cargo compartment, upper portion of the passenger cabin ....... broken into numerous segments, pieces of fractured cabin floor beams, and forward cargo compartment door which was snapped off"............................................... ...........................................................
North-South Order Including the Belly Structure
The actual sequence (Report and Appendices) in north-south order, with the stretch from km. 10 to 15 included, and with details as to location and altitude, is the following.
(1) The 'forward section of the fuselage ....... forward cargo compartment, upper portion of the passenger cabin ..... broken into numerous segments, pieces of fractured cabin floor beams, and forward cargo compartment door which was snapped off" fell at the stretch from km. 1 to 10 after separating from the aircraft at about 22,000 to 9,000 ft.
(2) The "nose section that included the entire cockpit, electronics equipment compartment, nose gear and both left and right cabin doors ......struck the ground" at about km.10 after separating from the aircraft at about 9,000 ft.
(3) The belly structure fell at the stretch from km., 10 to 15 after separating from the aircraft below 9,000 ft.
(4) The aircraft "midsection of fuselage, empennage, wings, engines and main landing gears ...... fallen in one piece" at km. 15.
Picture of Separation Sequence
Attachment IV is a picture of the separation sequence, With additional details, the following sequence emerges.
(1) The aircraft decompressed near km. 1 at about 22,000 ft. The upper part of the forward passenger cabin separated from the aircraft, whereupon pressurization from the cargo compartment below pushed the floor panels upward. In turn, the upwardly pushed floor panels (attached to the floor beam upper
flanges) pulled the upper flanges upward. Some upper flanges tore from the beam web together with the floor panels and seat tracks. The floor beams broke in the middle, nine upward and two downward. Some beams also broke at a second place near the left sidewall. Separating from the aircraft as it descended from 22,000 to 9,000 ft., the forward fuselage upper parts, floor panels and beams, and compartment sidewalls and door fell along the flight path at the stretch from km. 1 to 10. Bodies of 37 persons, mostly forward passengers, were found here.
(2) At km. 10 at about 9,000 ft., the nose section (including the cockpit, galley door, and entry door) tipped downward and sheared off at BS 360 bottom. It separated from the aircraft, carrying with it the electronics equipment compartment at BS 360 to 380 under S-17 floor level. Bodies of the pilot and co-pilot were here.
(3) The belly structure fell at the stretch from km. 10 to 15 after separating from the aircraft below 9,000 ft.
(4) The main fuselage, wings, etc. hit the ground at km.15. Bodies of 71 persons were here.
Far Eastern Floor Beams and Upper Flanges
It is unfortunate that Aloha Exhibit 15D omits the English Report appendices, because the descriptions of the floor beams and upper flanges at Appendix 11 (par. B18) and the Report itself (p. 17 and 25) are different. Appendix 11 (not stating the particular breaking direction of any beams) states that the upper flanges of "most" beams were subjected to an upward pulling force. Differently the Report (not stating that the upper flanges of "most" beams were subjected to an upward pulling force) variously states that "some" beams bent or broke in an upward direction.
Upward directional breaking can involve crack proportion downward from the upper flange in tension (lower flange in compression), which in turn can indicate upper lobe decompression (upward force from pressurization below). Conversely, downward directional breaking can involve crack propagation upward from the lower flange in tension (upper flange in compression), Which in turn can indicate lower lobe decompression (downward force from pressurization above.
The Appendix description (upper flanges of "most" beams were subjected to an upward pulling force) can only indicate upper lobe decompression (upward force
from pressurization below). However, the Report (Report only) description ("some beams bent or broke in an upward direction) is ambiguous and misleading. Capt. Leppard takes it to indicate upper lobe decompression ("most" upward and a "few" downward, whereas Mr. Harradine refutes that it indicates lower lobe decompression (a "few" upward and "most" downward. Who is correct? How many are "some" .
The report (Report only) does not tell us, but Appendix 11 tells us that Capt. Leppard is correct. "Most" upper flanges were subjected to an upward pulling force, meaning that "most" beams broke in an upward direction. The "few" that not inconsistently broke downward broke that way from the breakup force itself.
Aloha Transcript p. 831. Appendix 11 was unavailable. Mr. Swift, not knowing that "most" upper flanges were subjected to an upward pulling force (and not questioning the prevailing view), explains why "some" (a "few") upward directional breaking need not be inconsistent with lower lobe decompression. (By the same token, see below, "some (a "few") downward directional breaking need not be inconsistent with upper lobe decompression.)
There are 11 floor beams at BS 360 to 520. Their locations are BS 360, 380, 400, 420, 440, 460, 500, 500A, 500B, and 520. (The 737-100 does not have BS 500A and 500B)
All floor beams broke near the middle, and all beams (except the left halves at BS 360, 480, and 520) were recovered. The beams at BS 440, 460, and 500B (left halves) were selected for analysis, and Boeing reported (Appendix 12) that all three of them showed upward crack propagation (from lower flanges in tension to upper flanges in compression). At BS 440 and 460 the upward crack propagation was at the middle fracture, whereas at BS 500B it was only at a second fracture near the left sidewall. In sum, at the middle fracture, 9 out of 11 ("most") beams broke upward and 2 out of 11 (a "few") broke downward. The 9 out of 11 broke from pressurization force from below, the 2 out of 11 (like the left sidewell fracture at BS 500B) from breakup force.
Aloha Transcript p. 492. In reply to Chairman Lauber, Mr. Harradine explains that the Far Eastern floor beams showed upper flange compression. It is unclear whether he means "all", "most". "some", or a "few", but the preceding paragraphs show that only the "few" at BS 440 and 460 (middle fracture) and BS 500A (second fracture near left sidewall) showed upper flange compression. These "few" are the exception breakup force), not the rule. His explanation does not pass muster.
Aloha Exhibit 7A. This describes the condition of the Aloha floor beams at BS 360 to 520. The beams at BS 360 to 380 (forward end) and BS 500B and 520 (aft end) were not broken or cracked. Between these ends, five of the seven at BS 400 and 500B were broken and two were cracked nearly all the way through. Do the upper flanges show subjection to an upward pulling force? Because the aircraft did not break apart, probably no beams show upper flange compression (upward crack propagation from lower flange in tension to upper flange in compression. Do any?
Attachment III collects the description of the Far Eastern and Aloha floor beam and adjacent structure. It indicates similar distress patterns at the floor area, including left-side frame that broke just below the place where the beams join the frames. The Aloha description indicates frame breaking only at this place, whereas the Far Eastern description (for details see Appendix 11) indicates frame breaking at multiple locations including this place. Can this difference help explain why the Far Eastern aircraft broke apart whereas the Aloha aircraft did not?
When the aloha aircraft landed, how close was it to actually breaking apart?
If the Aloha aircraft had broken apart before landing, would the breakup mode have been similar to that of the Far Eastern aircraft? (After the upper lobe, floor beams, and sidewalls separated, the nose section tipped downward, sheared off at BS 360 bottom, and fell to the ground, carrying with it the electronics equipment compartment at BS 360 to 380 under S-17 floor level.)
Far Eastern Decompression Initiation Locations
I look at Chinese Appendix 19 and see the Far Eastern aircraft momentarily (km. 1 to 9) having a profile like that of the Aloha aircraft when it landed. I also see S-4L, S-14L, S-19L, and BS 360 as candidate locations for skin tearing, leading to the decompression, to initiate. I have tended to prefer S-19L, namely, at upward skin tear starting from S-19L (at the lower lobe) and reaching critical length above S-17 floor level (at the upper lobe). giving upper lobe decompression.
Aloha Transcript p. 826. I note Mr. Swift's preference for Aloha S-10L, over S-4L, which is similar to my preference for Far Eastern S-14L or S-19L over S-4L. Is There anything to favor or disfavor Aloha S-14L over S-4L or S-10L?
Aloha Transcript p. 845. I note the exchange by Chairman Lauber and Mr. Swift, and am intrigued by Mr. Swift's apparent perplexity over BS 360. This location has long perplexed me too. I now want to give more serious consideration to it as the probable initiation location for the decompression.
At one place the English Report (p. 7) states that the separated forward area (nose section etc.) includes the right-side galley door (BS 294 to 328), the left-side entry door (BS 321 to 344), and the electronics equipment compartment at BS 360 to 380 under S-17 floor level. Elsewhere, however, the report (p. 11) states that the forward area (nose section etc.) separated from the aircraft at BS 300 (center of the galley door). The Chinese text likewise reads BS 300.
However, I now wonder whether BS 300 may be a typographical error (for BS 360) in the original English that the Taiwanese failed to catch when rendering the English into Chinese. Is there anything to favor BS 300 (center of the galley door) over BS 360 (section splice and circumferential skin lap)?
If there is nothing to favor BS 300 over BS 360 as the decompression initiation location for both the Far Eastern and Aloha aircraft. The Notice of Proposed Rulemaking (NPRM of
May 24, 19??) at Docket No. ??-???-67-AD (53 Federal Register 190-1????) states that there have been numerous problems at BS 360. Chinese Appendix 19 also shows considerable skin fragmentation at this location.
Aloha Transcript p. ?20. Mr. Swift describes the Aloha aircraft as not badly corroded. Similarly, Appendix 11 (par. D.17) describes the Far Eastern belly structure as being in fairly good condition corrosion-wise. These descriptions suggest that our far Eastern/Aloha/early 737-200 problem may not entirely be one of corrosion or delimination on aircraft operated in a hot humid ocean climate, but may be partly be a problem of accelerated fatigue on understrength aircraft operated in a high cycle environment. Design differences from the 737-100 and 737-200C also may play a role. (The -100 is the original 737 design, the -200 is a stretched -100, and the -200c is a strengthened -200.)
737-200 line numbers 151 and 152. These are line numbers of the Far Eastern and Aloha accident aircraft, respectively.
737-200 line number 168. this was a -100 operated by China Airlines (before that by All Nippon) that crashed into the Taiwan Straits in 1986. The popular explanation is that it crashed when the pilot lost control after aborting an unsuccessful landing attempt at night. However, the Aloha Accident now opens such easy explanations to question. (Suppose Capt. Schornstheimer had unsuccessfully attempted to land and the aircraft had crashed into the sea.) Britain worked hard to recover the Comet that crashed into the Mediterranean near Elba in 1954. Recently, Italy recovered the DC-9 that crashed into the Tyrrhanian Sea in 1980. There are other examples of recoveries to assist important investigations. If hard evidence cannot satisfactorily explain why this China Airlines -200 crashed as it did perhaps we should request Taiwan to raise the aircraft.
737-200 line numbers 519, 591, 706, 896, 1033, and 1241. These are -200s presently operated by Southwest Air Lines of Japan (Okinawa) in hot humid ocean NTSB, September 20, 1988, page 11
climate. They are only two to 10 years old, but they also are high-cycle operation. Aviation Week (August 1, 1988) has reported that Southwest is seeking to replace all of these -200s (its entire 737 fleet) with 737-400s. Perhaps we should request Japan to inspect and report on the structural condition of those Southwest -200s.
737-100s. Continental airlines presently operates 17 early -100s having line numbers between 9 and 127. They are 19 or 20 years old. They represent more than half the entire -100 fleet (only 30 were produced). Their present structural condition should be fairly representative of the age-20 structural conditions for the original 737 design. results of inspecting these Continental -100s may provide a useful standard by which to evaluate the present structural condition of early -200s.
737-200Cs. The Aloha Transcript describes Aloha N4906, having line number 173, as the most highly corroded 737 in existence. from this, one would think that it is a -200 that has spent its entire operating life in hot humid oceanic Hawaii. However, until 1986, it was operated by Wien and Markair in cold northern climates. It also is a 737-200C. Perhaps we should inspect other early -200Cs to determine whether there may be a pattern of early -200Cs, including corroded -200Cs, showing greater breakup resistence than early -200s.
Accurate evaluation of the role, if any, of the understrength /accelerated fatigue factor is essential to recommending specific inspection and modification measures needed to face up to our Far Eastern/Aloha/early 737-200 problem. The suggested collateral inspection may give some helpful insights. More corrosion control alone may not be the full answer.
Comets never operated here, but early 737-200s were and are operated predominantly in this country. For this reason, the Far Eastern accident was and is as important to Americans as Foreigners. The investigation was and is an important
as the Aloha investigation. Yet, and although we finance participation in foreign accident investigations primarily to learn for our own benefit, we somehow failed to take full advantage of the learning opportunity offered by the Far Eastern accident.
Some will blame the NTSB for failing to remain more directly involved is investigating and reporting such an important accident. Others will blame the FAA for failing to issue the SB 1039 Airworthiness Directive in 1982, when it issued the SB 1042 AD shortly after the accident. SB 1042 recommends concurrent effectuation of the SB 1039 service Bulletin, but the FAA acted only on SB 1042. Even that limited action was deficient, see NPRM of May 23, 1988. I personally place much of the blame on our current legislative system which so discourages full and explicit reporting of accidents. Someday we must face up to this aspect of our aviation safety problem.
The odds favor re-occurrence of the Far Eastern/Aloha/early 737-200 accident with in the next few years, partly because of lost years and modification opportunities. We cannot prohibit repeat accidents. We can only try to prevent them and thereby limit their number. I trust that my comments will be helpful. Learning from two accidents can be much more fruitful than learning from just one.
Yours Truly, /s/ Norman Jensen 1543 Veteran Avenue Los Angeles, California 90024