PBl9-910404

.--~

NATIONAL TRANSPORTATION SAFETY BOARD

WASHINGTON, D.C. 20594

AIRCRAFT ACCIDENT REPORT

ALOHA AlRUNES, FLIGHT 243 BOEING 137-200, N7371', NEAR MAUl, HAWAII APRIL 28, 1988

NTSBlAAR-89I03

UNITED STATES GOVERNMENT

1- Report No. NTSBIAAR-89/03

'12.- PB89- 91 0404

Government

TECHNICAL REPORT DOCUMENTATION

.

Accession No .

~----.

.,AGE

3. Recipient's Catalog No.

4. Title and Subtitle

Aircraft Accident Report--Aloha Airlines, Flight 243, Boeing 737-200, N7371 t, near Maui, Hawaii, April 28. t988 , 7. Author(s)

5. R4!!)OrtDate
June 14, 1989

6. Performing Organization
Code 8. Performing Organization Report No.

I\
I

,
9 Performing Organization Name and Address National Transportatu:)n Safety Board Bureau of Accident Inv2St,gation Washington, D.C. 20594

1.

10. Work Unit No. 4886A 11. Contract

Of'

Gram No.

l
I I
I

13. Typ4!of Repon and Period Covered Aircraft Accident Report April 28, 1988 14. Spo~sorin1 Agencr Code

112. Sponsoring Agenc~ Name and Address NA IIONAl TRANSPORTATIONSAFETYBOARD Washington, D.C. 20594 Notes

15. Supplementary

16. Abstract: On April 28,1988, at 1346, a Boeing 737-200, N73711, operated by Aloha Airlines Inc., as flight 243, experienced an exp'osive decompression and structural failure at 24,000 feet. while en route from Hila, to Honolulu, Hawaii. Approximately 18 feet of the cabin Skin and structure aft of the cabin entrance door and above the passenger floorline separated from the airplane during flight. There were 89 passengers and 6 crewmembers on board. One flight attendant was swept overbOard during the cecomor~sion and is ~resurr.ed to haVE- been fat;lIy injured; 7 passengers and 1 flight attendant received serious injurie!. The flightuew performed an emergency descent and landing at Kahului Airport on the Island of MauL

The safety issues raised in ~iS report include: the quality of ai:- cartier In;)intenance programs and the FAA surveillance of those programs, the engineering airworthiness of the 8-73-7 with particular emphasis on multiple site fatigue cracking of the fuselage lap joints. the human factors aspectS of air carrier maintenance and inspection for the continuing ~irworthiness of transport category airplanes. to include repair procedures and the training. certification and .Qualificatjon of mechanics and inspectC)f"S.

t7. Key Words

decompression; disbonding; fatigue cra~king; corrosion; multiple site damage (MSD); FAA surveillance; maintenance ·program; nondestructive inspection

18. Distribution Statement

This document is available to

the P'-Iblic through the NatioNlI

Technical Inform~tion Service. Springfield, Virginia 22161 21. No. of Pages 22. Price

19. Security Oassifkation

(of this report) UNCLASSIFIED

20. Security Oassification (of this page) UNCLASSIFIED

262

NTSB Form 1765.2 (Rev. 5/81~)

CONT~NTS

EXECUTIVE SUMMARY

v
1

1.

1.1 1.2

1.3 1.3.1 1.3.2 1.3.4

1.4 1.5 1.3.3

1.6 1.6.1 1.6.2 1.6.3 1.6.3.1 1.~.3.2 1.6.3.3 1.6·.3.4

1.7 1.8

1.9

1.10 1.11 1.12 1.13 1.15

1.14

1.1S.1 1.15.2

1.15.3 1.15.4 1.16 1.16.1 1.15.2

1.16.3

1.17.3 1.17.4 1.17.5

1.17 1.17.1 1.17.2

1.17.6

1.17.7 1.17.8 1.17.9 2.1 2.2

Hinory of the Flight , . Injuri~ to Persons ..•....................... , •.......•.....•.••. Damage to Airplane •.........•.. : ...................•..•.....•. Genera. . . Fuselage Separation Area : .. Additional Airplane Damage .........................•......... Pressurizatio:1 System .....................................••... Other Damaae ..................•...................•......... Personnel Information _ _. _.. _. _.. _ . Airplane !nformation .....•..............•..........••....•...• General .. _..••...•.................................•......... lap Joint Design and Bonding History ........•.•.....•.......... Aloha Maintenance History . Maintenance Prog;am _.. _. _.. .. Maintenance Records Review " .. Service Bu lIetins . FAA Airworthiness Directive (AD) Compliance AD 87-21-08 . Meteorolo~icaf Information ...•...............................• Aids to N~vl~ation ..•.....•......................... CommunIcations . Aerodrome Information .........•............................. flight Recorders , _. _ _........•. _.. Wreckage and Impact Information _ _ . Medital and Pathological Information ....................•...... Fire .•.•••••••..•.••............•............................. S~rvival Aspects .•...••.•.•.•.••....•..•........•...••.....•... Supplemental Oxygen S)'Stems ' . Sea Search •.•.. ......•................................... _ . Rescue and Firefighting Response _ _ . Ambulance Response •.•.••••••..•.•..•.•........•....•.•....•. Tests and Research ....•...•.•.••...••..•...•.......•.••.......• Pressurization System .. _ . Eddy Current and Visuai :nspection ...................•.......... Materia's Laboratory Examination . Additionallnformation ...•.••••.••.••••..•.•... _.............• Genera. Inspection of Other Aloha Airlines Airplanes ...........•. The 8-737 Fail-Safe Design ..•................................... tn·SerJice Model ~uselage Tests ...•....... __ . service Difficulty Report Information _..•..•. _....•....•.......•. Supplemental Structural Inspection Program (SSIP) •.••••.•....••. FAA Surveillance of Aloha Airlines Maintenance ...•.. _.•.•....... Boeing Commerdal Airplanes Customer VISits •.........•......... The National Aviation safety Inspection Program ..•............•• Subsequent FAA Action .....................•.•................
< ••••••••••

fACTUAL INFORMAnON

..

~,

10

5 5 5 8

11 11

11 12 13 21 21 24 26

12

23

25
26

27
27

27 27 • 27 28 • 28 28 28 29 29 29

29
29 30

31
31 34

35
36 36 37 40 4S

42

2.

2.3
2A

ANALYSIS General •. . . .• .•..•..........•.•....•........•..•..........•• 47 Origin of Fuselage Separation •• • • • . • . • . • . • . • • • • • • . . . . • • . . . . • • . • 4'1 Fuselage Separation sequencE •••••••..•..•.•...•.....•... _.. _. _ SO Aloha Airlines Maintenance Pr~ram ...........••..........••.•. 51

iii

"tiift

---~-"~

-if·
i!

2.4.' 2.4.2 2.S 2.5.1 2.5.2 2.5.3 2.6

2.4.3

Effectil;eness of Inspections . Aloha Airlines Corrosion Control

2.6.1 2.7
3.

2.6.2 2.6.3

F~.A Responsibilities .............•.........•.•......•.•.....••. Issuance and Claritv of Airworthiness Directives .....••...•....... Needed Research Corrosion Control and NOI FAA.Oversight Boeang Boeing 737 Certification .....................................•. Boeing Structures Classification ............•.................... Boeing Visits to Aloha Airlines ...................•..........••..

EngineeringServices

.............•....•.....•.•••.••

.
. . .

on

62

53 58 S9 61 61

,~
,

'1,. •.•••

\
\.

Operational Considerations
CONCLUSIONS Findings •.......... Probable Cause RECOMMENDATIONS
+ •••

+.

63 67 67 68 68 69
71

3.2 4. 5.1

3.1

•••••••••••••••••••••

•••••••••

•••••

•••••

73
74
~.

APPENDIXES Appendix A-Investigation and Hearing ........•................. Appendix B-Personnellnformation . Appendix C--Boeing Service Bulletin 737-53-1039, Rev. 2 . -FAA Airworthiness Directive 87-21-08 •..•••.•..••.. -Boefng Service Bulletin 737-53A 1039, Rev. 3 . -Boeing Nondestructive Test. 737-53-30-03 . Appendix 0-8oeing Service Bulletin 137-53-1076 '" . Appendix E-SpeciaJist's Factual Report of Investigation, CVR Flight Data Readout Report of Investigation ..•••.......•....••.. Appendix F-Afoha Airlines Non-Destruct~ve Testing Reports . Appendix G-Summary of Previous Repairt on N7371 1 ..•.....•... Appendix H-SOR Summary •.••.•...•...••..•...............•.. Appendix I--Boeing MGOS Airlines Maintenance Evaluation . Appendix J-FAA Report on Multiple Site Cracking, December 18, 1986 ..............................•.......

79 80

112 115
151

82

156

:,

157 191 208 210

214 245

iv

,
EXECUTIVE SUMMARY

On April 28~ 1988 at 1346, a Boeing 737-200, N73711, operated by Aloha Airl ines Inc., as flight 243 experienced an explosive decompression and structural failure at 24,000 feet, while en route from Hilo, to Honolulu, »awaii. Appro~imately 18 feet from the cabin skin and structure aft of the cabin entrance door and above the passenger floorline separated from the airplane during flight. There were 89 passengers and 6 cret1l1tembers board. on One flight attendant was swept overboard during the decompression and is presumed to have been fatally injured; 7 passP~gers and I flight attendant received serious injuries. The flight crew performed an emergency descent and landing at Kahului Airport on the Island of Mau;.
J t

The Natiana1 Tra;:'sportat Safety Board determines that the i on probable cause of this accident ~as the failure of the Aloha Airlines maintenance program to detect the presence of significant disbanding and fatigue damage which ultimately led to failure of the lap joint at S-10l and the separation of the fuselage upper lobe. Contributing to the accident were the failure of Aloha Airlines management to supervise properly its maintenance force; the failure of the FAA to evaluate properly the Aloha Airlines maintenace program and to assess the airline's inspection and quality control deficiencies; the failure of the FAA to require Airworthiness Directive 87-21-08 inspection of all the lap joints proposed by Boeing Alert servtce Bulletin SB 737-53AI039; and the lack of a complete terminating action (neither generated by Boeing nor required by the FAA) after the discovery of early production difficulties in the 8-737 cold bond lap jOint which resulted in low bc~d durability, corrosion, and premature fatigue cracking. The safety 'issuesraised in this report include: o a The Quality of air carrier maintenance programs and the FAA surveillance of those programs. The engineering design, certification, and continuing airworthiness of the B-737 with part.icular emphasis on multiple site fatigue cracking of the fuselage lap jOints. The human factors aspects of air carrier maintenance and inspect ion for the cont i nui"g a i rworthiness of transport category airplanes, to include repair procedures and the t~aining, certification and Qualification of mechanics and inspectors.

Recommendations concerning these issues were addressed to the Federal Aviation Administration, Aloha Airlines, and the Air Transport Association.

NATIONAL TRANSPORTAnOM VASHIRGTOH. D.C.

SAFETY BOMD 20594

I
1.1

AIRCRAFT ACCIDENT REPORT ALOHAAIRLINES, fLISHT 243

BOEING 737 -ZOO, 1173711,

MEAl MIl.

APRIL 28, 19sa

HAWAn

1.

FACTUAl INFORMATION

History of the Flight

On April 28~ 1988, an Aloha Airlines Boeing 7377 N73711~ based at the Honolulu International Airport, Hawaii, was scheduled for a series of interisland flights to be conducted under Title 14 Code of Fed~ral Regulations (CfR) Part 121. A captain and first officer were assigned for the first six flights of the day with a planned first officer change to complete the remainder of the da,ily schedule. The first officer checked in with the dispatch office about 0500 Hawaiian standard time at the Aloha Airlines Operations Facility. After fami1iarizing himself with the flight operations paperwork, he proceeded to the Aloha Air1 tnes parking apron and perfonaed the preflight inspection required by company procedures before the first flight of the day. He stated that the airplane maintenance log release was signed and that there were ~ open discrepancies. He prepared the cockpit for the external portion of the preflight, exited the airplane in predawn darkness, and performed the visual exterior inspection on the lighted apron. He stated that he found nothingunusual and was satisfied that the airplane was ready for flight. The captain checked in for duty about 0510; he completed his predeparture duties in the dispatch offlce and then proceeded to the airplane. " The crew flew three roundtrip f11ghts~ one each from Honolulu to and Kauai. They reported that all six fHghts were uneventful and that all airplane systems performed in the normal and expected _nner. Flightcrew visual exterior if!'spections between fl ights were not required by Federal Aviation Admi~istration (fAA) accepted compaoy procedures7 and none were performed.
Ht lo, Maui,

At 1100, a scheduled first officer change took place for the remainder of the day. The cr,ewflew from Honolulu to Maui and then frOllMaui to Hi10. As with the previous flights of the day~ no system, powerplant, or structural abnormalities were noted during these operations, and the flights were uneventful. Neither pilot left the airplane on arrival in Hilo~ and the crew did not perform any visual exterior inspection nor were they required to do so .

At 1325, flight Z43 departed Hila Airport en route to Honolulu as part of the normal scheduled service. In addition to the two pilots, there were three flight attendants, an FAA air traffic controller, who was seated in the observer seat in the cockpit, and 89 passengers on board. Passenger boarding, engine start, taxi. and takeoff were uneventful. The planned routing for Alcha flight 243 was from Hilo to Honolulu at flight level 240. Mau; was listed as the alternate landing airport. The fi rst off; cer conducted the takeoff and en route c1~mb from Hi10. The captain performed the nonflying pilot duties. The first officer did not recall using the autopilot. The flight was conducted in visual meteorological conditions. There were no advisories for significant meteorological information (SIGMET) or airman's meteorological information (AIRMET) valid for the area along the planned route of flight. No unusual occurrences were noted by either crewmember during the departure and climbout. As the airplane leveled at 24,000 feet, botp pilots heard a loud "clap" or "whooshing" sound followed by a wind noise behind them. The first off; cer s head was j erked backward, and she stated that debris, including pieces of gray insulation, was floating in the cockpit. The captain observed that the cockpit entry door was missing and that "there was blue sky where the first-class ceiling had been." The captain immediately tool< over the controls of the airplane. He described the airplane attitude as roll ing slightly left and right and that the fl ight controls felt "loose."
I

.-; ,..

Because of the decompression, both pilots and the air traffic controller in the observer seat donned their oxygen m~sks. The captain began an emergency descent. He stated that he extended the speed brakes and descended at an ;ndicated a;rspeed (lAS) of 280 to 290 knots. Because of ambient noise, the pilots initially used hand signals to cOlllllunicate.The first officer stated that she observed a rate of descent of 4,100 feet per minute at some point during the emergency descent. The captain also stated that he actuated the passenger oxygen switch. The passenger oxygen manual tee handle was-not actuated. Wh~n the decompression occurred, all the passengers were seated and the seat belt sign was illuminated. The No. 1 flight attendant reportedly was standing at seat row 5. According to passenger observations, the flight attendant was immediately swept out of the cabin through a hole in the left side of the fuselage. The No.2 flight attendant, standing by row 15/16, was thrown to the floor and sustained minor bruises. She was subsequently able to crawl up and down the aisle to render assistance and calm the passengers. The No. 3 f1 ight attendant, standing at ~ow 2, was strUCK in the head by debris and thrown to the floor. She suffered serious injuries including a concussion and severe head lacerations.

The first officer said she tuned the transponder to emergency code attempted to notify Honolulu Air Route Traffic Control Center (ARTee) that the flight was diverting to Maui. Because of the cockpit noise level, she could not hear a:l), radio tr-ansnt ss tons , and she was not sure if th~ Honolulu ARTCC heard the communication.
7700 and

Although Honolulu ARTee did not receive the first officer's initial communication} the controller working flight 243 observed an emergency code 7700 transponder }'eturn about 23 nautical miles (nmi) south-southeast of the Kahalu; Airport, Maui. Starting at 1348:15. the controller attempted to communicate with the flight several times without success. When the airplane descended through 14,000 feet, the first officer switched the radio to the Maui Tower frequency. At 1348:35, she informed the tower of the rapid decompression, declared an emergency, and stated the need for emergency equipment. Maui Tower acknowl edged and began emergency notifications based on the first officer's report of decompression. At the local controller's direction, the special i st working the Mau; Tower clearance delivery position notified the airport's rescue and firefighting personnel, via the direct hot 1ine, that a B-737 had declared an emergency, was inbound and that the nature of the emergency was a decompression. Rescue vehicles took up alert positions along the left side of the runway.

At the Maui Airport, ambulance service was available from the nearby community when notified by control tower personnel through the local "911 telephone number. Tower personnel did not consider it necessary at that time to call for an ambulance based on their understanding of the nature of the emergency_
II

At 1349:00, emergency coordination began between Hono'lulu Center and Mau; Approach Control. Honolulu advised Maui Approach Control that they had received an emergency code 7700 transponder return that could be an Aloha 737 and stated, "You might be prepared in· case he heads your way." Maui Approach Control then advised Honolulu Center that flight 243 was diVerting to land at Maui. The local controller instructed flight 243 to change to the Maui Sector transponder code to identify the flight and indicate to surrounding air traffic control (ATC) facilities that the flight was being handled by the HaUl ATC facility. The first officer changed the transponder as requested. The flight was operating beyond the local controller's area of radar authority of about 13 nmi. At 1350:58, the local controller requested the flight to switch to 119.5 MHz. (approach frequency) so that the approach controller could monitor the flight. Although the request was acknowledged~ the flight was not heard on 119.5 MHz. Flight 243 continued to transmit on the local controller frequency.

At 1353:44, the first officer informed the local controller, "We're going to need assistance. We cannot communicate with the flight attendants. We'll need assistance for the passengers when we land." An ambulance request was not initiated as a result of this radio call. The first officer also provided the local controller with the flight's passenger count, but she did not indicate the fuel load. The local controller did n6t repeat the request for the fuel load even after a query from the chief of the emergency response team. The captain stated that he began slowing the airplane as the flight approached 10,000 feet mean sea level (msl). This maneuver is required as a routine operations practice to comply with ATC speed limitations. He retracted the speed brakes, removed his oxygen mask, and began a gradual tUrn toward Hau;'s runway 02. At 210 knots lAS, the f1ightcrew could communicate verbally. The captain gave the cotmland to lower the flaps. Initially flaps 1 were selected, then flaps 5. When attempting to extend beyond flaps 5, the airplane became less controllable, and the captain decided to return to flaps 5 for the landing. Because the captain found the airplane becoming less controllable below 170 knots lAS, he elected to use 170 knots lAS for the approach and landing. Using the pu~lic address (PA) system and on-board interphonet the first officer attempted to communicate with the flight attendants; however, -there was no response.
At the command of the cepta+n, the first officer lowered the ,landing gear at the normal point rn the approach pattern. The main gear indicated down and locked; however, the nose gear position indicator light did not illuminate. Hanua1 nose gear extension was selected and still the green indicator 1 ight did not illuminate; however, the red landing gear unsafe tndtcatov light was not illuminated. After another manual attempt, the handte was pl aced down to complete the manual gear extension procedure.

The captain

because the center jumpseat was occupied and the captain believed it was urgent to land the airplane immecliately. At 1355:05, the first officer advised the tower, ·We won't have a nose gear," and at 1356:14, the crew advised the tower, ·We'll n~ed all the eQuipment you've got.n
capt.a+n sensed a ya.,.ing motion and determined that

said no attempt was made to use the nose gear downlock viewer

While advancing the power levers to maneuver for the approach, the
the No.1· (left) engine

had failed. At 170 to 200 knots lAS, he placed the No. 1 engine start switch to the "flight" pos+t+on in an attelmpt to start the engine; there was no response. A normal descen~ profile was esttblished 4 miles out on the final approach. The captain said that the airplane was "shaking a little, rocking slightly and felt springy."

Flight 243 landed on runway 02 at "aui's Kahului Airport at 1358:45. The captain said that he was able to make a normal touchdown and landing rollout. He used the No.2 engine thrust reverser and brakes to stop the airplane. During the latter part of the rollout, the flaps were extp.nded to 400 as required for an evacuation. An emergency evacuation was then accomplished on the runway. After the accident, a passenger stated that as she was boarding the airplane through the jet bridge at Hilo, she observed a longitudinal fuselage crack. The crack was in the upper row of rivets along the S-10L lap jOint, about halfway between the cabin door and the edge of the jet br'idge hood. She made no mention of the observation to the airline ground personnel or flightcrew. 1.2 Injuries to Persons Injuries Fatal Serious Minor None Total Crew
1 0

Passengers
0

Others
0

I"

a 5

57 II
89

1**
1

1*

zs 95

57

*lost in flight; a sea search was unsuccessful. **Air traffic controller seated in the observer seat in the cockpit .
1.3

Damage to Airplane Genera'

1.3.1

A major portion of the upper crown skin and structure of section 43 separated in f1ight causing an explosive decompresslon1 of the cabin. (See figures 1 and 2.) The damaged area extended from slight1y aft of the main cabin entrance door, rearward about 18 feet to the arpa just forward ~f the wi ngs and from the 1eft side of the cabi n at the floor 1eve1 to thp. right side window level. The value of the airplane was estimated at about $5 million. As a result of the accident, the airpla:'lewas detennined to be damaged beyond repair. It was dismantled on the site and sold for parts and scrap.

1·ElCPlosivlt deCOMpression" in this CIS. indicltes' and hof.a fro. c~bin air rele ••• d under pr ••• ure rather ch •• fcat explosive device.

I violent expansion than the effetts of a

MISSING AREA
BODY
STA 259

BODY
STA 360

BODY
STA 727

seCT

SECT

SECT

SECT

Figure l.-·Boeing 737-200·-Body Stations, Stringers, and Section Locations.

Figure Za.--General view, left sid~ of forwlrd fuselage, H737}1

I»

Figure 2b.--General view. right side of forward fuselage, N73711. Arrow lIar-Ks fragmer,ts of S-4R lodged in the leading edge fllp~

1.3.2

Fuselage Separation Area •

The fuselage structure cons ists primarily of skin, frames,z and stringers.! Skin panels are joined longitudinally at lap jOints where tr.e sheet &etal of the upper skin panel overlaps the sheet metal of the lower skin panel about 3 inches. When manufactured, this overlapped area was bonded and riveted tot} th tll~ee rows of countersunk rivets. (See 1. 6.2 lap Joint Design and Bonding H~story.) The area ."nere the structure was missing extended from body sta!i0n4 (BS) 360 aft to about BS 540, a~d circumferentially from just above the floor on the left side of the airplane (at $- 1Sl), across the crown and dOW'n the rig.ht side to a position above the window belt (at S··lOR). The structure from the top of the window belt to the fl60r on the right side was distorted severely and bent outward !DOre than 900• The skin had peeled in this area leaving th~ frames~ stringers: and window forgings in place. On the leit side below the floor level~ the skin had peel~ off the structure in large V-shaped areas. Five consecutive floor beallss at as 420. 440, 460, 480 and 500 were broken ~ll the way through. Also, the adjacent floor beams at BS 400 .nd SOOA were cracked nearly all the way through. The fractures and cracks were slightly to the left of the airplane centerline. The frames at these S~ seven stations were brokep on the left sid~ just below the floor beams. Most of the c'nter floor panels from BS 360 to BS 947 were displaced upward except in the o¥erwing area. The right side cabin floor panels had not been displaced and llttlf if iny distress had occurred at the fastener locations for these ~nels. However" on the left side 'Jf the airplane between as ~OO and as soc i.loag the inboard sea! tnck. there was extensive floor panel dtsplac. . nL . The floor panels ~c! dispiaced upward and had reached their IaXl~ dfspl.ceDent of 4 inches at as 440 (matching the displacement of the brote~ floor bea.s).
t

tr~ ~t~.

bet~J\

A f~$elage section f~

this was the oftl1 significlnt piece of structure from the damaged

tM leading edge

SS 365 to

n.p

.nd inboard side of the right engine

as

420 between S·4R and S-SR was

S,t~t~••,. It'
•f~~*.f~ ••
t~ •• ~

.~f,~. i.h.~tv.j~rl .·rwct~r.t •••b.~. ~f t~. fu•• l•••• ~~ ••• f,•• el. Clftt.,lJnt ~f ". top 4~ .~. ~~••l.~•• t eLI .~. t •• -t e') tattor, ••• $t •• U for•• rd

'*

i ••

......

1~.

".~.!"_,

~'.fi he f~ •••_.~ti.f •• •• ,••

tit "'._

t.

•,"'.Ii.",•.
... ~
,

,~~rf"~~.' ••ff~

4... ~

.... ft.~

til: ~".'. •• _ ~~«~.'J 'ro...••

,.'ftt r. ~...

r.~c. P.J~t "••r

.l._.
S ••

fl. f.""I'
t~.

"0"

••••

of tl•

ut.d

'!I ''''If#
:,)

~~~

". H.""
e,

"wl'oI'iJ,..

t,. tt ..-., l*-o,dtO{'l!t.n •••

.~P•• ft

r jt

e•

,!fl..

••••• ,.. for t.&ift ft'l!'

", floor • .t S·

t'.

area that was recovered. The recovered piece contained two skin repairs along S-4R. This section and several samples cut from the remaining fuselage ski n were submitted to the Safety soard s Materials Laboratory for further analysis. (See 1.16.3 Materials Lab~ratory Ana1ysis.)
I

An examinat!on of the remaining structure ilTJlllediately surrounding the separation area, including skin, rivet, and st rtnqer deformations, revealed the following failure patterns:
o
o

left side. BS 360 to 540--The skin was peeled from the structure in a down and aft direction.
R~ght side. BS 360 to 540--The skin was peeled from the s~ructure in a down and aft direction, changing to directly aft near BS 540.

Circumferential break at BS 360--Fracture of the stringers and deformed rivets indicated that the separated structure was pulled generally aft eA~ept between S-SL and S-4R, where the direction was about 300 to the right of directly aft. Fractures and deformations showed that the separated skin had generally pulled through the butt joint rivets~ except at several locations where the separation was in the butt spl ice strap. Circumferential break at BS 540. left side·-From the top center of the fuselage to S-lOL, the skin fracture transit ioned from several inches forward of BS 540 to about 20 inches forward of 8S 540 and was not associated with any rivet line. At the S-10L lap joint, the fracture fo11owed the upper rivet line of the skin lap joint from a position 20 inches forWard of BS 540 to a position about 6 inches forward of BS 540. There were indications of preexisting fatigue cracks associated with seven consecutive rivet holes along this portion of the rivet line. From S-lOl to the floor 1 tna, the skin generally had separated several inches forward of station 540. Cjrcumferential break at 8S 540. right side~-From the top center of the fuselage to S-IOR the fractures in the stringers and deformed rivets indicated that the separated structure was pulled directly forward. In the 'Vicinity of S-llR, a small area of structure had been pulled forward and up. Below S-llR. the skin had been torn but the departure direction was unclear.

Indicitions of preexisting cracks were found in the S·10L lap Joint forward of as 540, on each side of a rivet hole in the BS 360 butt strap near S·1R, and in lap jOint rivet holes in a piece recovered from the right wing.

10

All other fractures adjacent overstress separations.

to the separation

area were

typical

of

The fracture surfaces and the immediate areas surrounding the separation perimeter generally were cerrcs+en free. However, areas of corrosion and disbonded surfaces6 were noted in the bytt joints at BS 360 and 540. Additionally, some areas of Dulged skin were noted on the intact skin lap joints and circumferential butt jOints that remained ~ith the airplane. 1.3.3 Additional Airplane Damage There was minor impact damage on the leading edges of both wingst although the damage was more extensive on the right wing. In addition, both horizontal stabilizers and the lower portion of the vertical stabilizer had random dents in the leading edges. The inlet cowls of both engines were dented, and several first stage fan blades of both engines were damaged. Remnants of fuselage structure were found against the inlet guide vanes and embedded in the acoustic liner of the right engine. A cable in the closed loop cable system for the left engine thrust lever and a cable in the left engine start lever system were broken near a pulley cluster located in the leading edge of the left wing irrmediately inboard of the engine strut. The broken start lever cable prevented motion .of the fuel control to the start position; the broken thrust lever cable prevent~d any power increase on the engine. The left engine fuel control was found in the "cutoff" position. Initial examination of the broken cables .showed signs of heavy corrosion in the area of the separat ion. Routing of these cables between the cockpit and the left engine pod was traced through the area of maximum up~ard floor defection at BS 440 under the cabin floor. The cables were retained and submitted to the Safety Board's Materials laboratory for further examination. (See 1.16.3 Materials laboratory Analysis.) The upper fuselage crown separation resulted in damage to overhead wire bur~les, and a number of circuit breakers in the cockpit were tripped. Most of ~hese ctrce tt breakers were related to passenger service un+t and lavatory tliiring. The potable water line was leaking and its conduit was broken. The pitot line and the static line to the flight data rEcorder (FOR) were broken, as was the conditioned air distribution dueting. The passenger oxygen manifold was severed which prevented use of the passenger oxygen system; however, the f1ightcrew oxygen system was undamaged. The flightcrew and passenger oxygen cylinders "ere fully discharged. Both engine fire bottles were emp~y~ and both of the engine fire extinguisher switches in the cockpit had been i~tivated, per the airplane emergency evacuation procedures.

60f.boftded

iftdicheS

the

.eparation

to,.tber)

lurfacel;

In this ca.e. alu.inu.

of previo\l8ly joined fUlela,. Ikin panele.

(glued

....

11

The hydrauli c system was not damaged. A11 the 1anding gear were down and locked, the flaps and leading edge devices were fully extended, the spoil ers were retracted, and there was no 1ass of hydraul ic flu id. An
examination

one of the two bulbs was burned out and that the module was slightly loose in

of the nose gear

position

indicator

1ight

module revealed

that

its housing. No other discrepancies were found tn the nose gear position indicating system. 1.3.4 Pressurization System

The main (aft) outflow valve and the forward outflow valve were fully closed. The forward outflow valve receives position signals from the main outflow valve. The pressurization controller was found in "automatic" and the flight/ground mode selector switch was found in the "flight" position. The flight position causes the cabin altitude contro11er to conform to the selected flight profile and also to modulate ths main outflow valve toward the closed position to pressurize the cabin slightly (0.1 psi) during ground operation. The switch is normally set to flight after engine st.art to pressurize the airplane; the switch ;s set to "ground" to depressurize after the landing rollout. Continuity checks showed normal system operation. All relevant system components were removed from the airplane for further funct lonal tests. (See 1.16.1 Pressurization System.)

1.4 1.5

other Oamage None. Personnel Information

The flightcrew consisted of the captain, first officer, and three flight attendants. (See appendix B.) The captain was hired by Aloha Airlines on May 31, 1977, as a 8-737 first officer. He was upgraded to captain on June 1, 1987. He possessed a current first-cl ass medical certificate with no 1imitations. He held an airline transport certificate with a type rating for the 8-737. At the time of the accident. the captain nad accrued about 8.500 total flight hours with 6,700 hours in the 8-737. HIS pilot-in-cammand time with Aloha Airlines was 400 hours, all in the 8-737. with a limitation for corrective lenses. She holds an iirline transport certificate without type ratings. At the time of the acctdent , the first officer had accrued about 8,000 total flying hours with about 3.500 hours in
the 8-737.

8-137 first officer. She possessed a current first-class medical tertificate

The first officer was hired

by

Aloha Airlines on June 4, 1979, as a

A dispatch records review indicated that the crew had complied with all relevant flightcrew duty time limitations.

12
Flightcrew training records included documentation of nonnal and emergency pr~cedures training. The Aloha Airlines flightcrew training program outline required emphasis on cockpit resource management (CRM) concepts; however, the training program did not include a specific CRM course, and line oriented flight training (lOFT) programs were not conducted, nor were they required by regulation.

1.6 1.6.1

Airplane Information General

The accident airplane, N73711, a Boeing 737-297, serial number 20209, was manufactured in 1969 as production Hne number 152. It was equipped with two Pratt and Whitney JT8D-9A engines. The airplane was delivered on Kay 10, 1969, to Aloha Airlines, the original operator. According to the limitations section of the FAA-approved Airplane Manual for B-737, N73711, the maximum zero fuel weight is 88~OOO pounds. the maximum certificated takeoff weight is 100,000 pounds. The actual weights for the departure on the accident flight were calculated at 80,253 pounds zero fuel weight and 93,133 pounds actual takeoff weight. The center of gravity (CG) computed for departure was 22 percent mean aerod~namic chord (MAC). The calculated CG limits for this gross weight were 4.0 percent and 30.5 percent MAC~ respectively.
Flight

The Aloha Airlines fleet consisted of eleven airplanes. all 8~737s. Four of the airplanes were considered high time, in excess of 60,000 cycles; onE was the worldwide fleet leader. At the time of the acctdent , the N73711 had accumulated 35t496 flight hours and 89~680 flight cy~les (landings), the second highest number of cycles in the worldwide B-737 fleet. Due to the short distance between destinations on some Aloha Airlines routes, the maximum pressure differential of 7.S psi was not reached on every flight. Therefore, the number of equivalent full pressurization cycles on the accident airplane is significantly less than the 89,680 cycles accumulated on the airplane. A review of 8-131 accidents and incidents reported to the Safety Board revealed one previous mishap involving N73711. On February 21. 1979, the airplane was operated into clear air turbulence that resulted in serious injury to two flight 3ttendants. No record of any damage or required repair to the airplane was found. There had been one previous accident involving in-flight structural failure of a B-737 fuselage. A Far Eastern Air Transport·, Ltd. (FEAT) e-737#200~ Republic of China registration B-2603 experienced an explosive
p

13

decompression and in-flight breakup on August 22, 1981.7 The accident occurred near Sanyi, Miaoli. Taiwan, and was investigated by the Civil Aeronautics Administration (CAA) of the Ministry of Communications, Taiwan, Republic of China. The Safety Board, Boeing, and the FAA participated ;n the investi,ation. The Republic of China CAA determined that the probable cause of the accident was: extensive corrosion damage in the 10\olerfuselage structures, and at a number of locations there were corrosion penetrated through pits, holes and cracks cue to intergranular corrosion and s~in thinning exfoliation corrOSion, and in addition, the possible existence of untietected cracks because of the great number of pressurization cycles of the aircraft (a total of 33,313 landings), interaction of these defects and the damage had so deteriorat~d that rapid fracture occurred at a certain flight altitude and pressure differential resulting rapid decompression and sudden break of p2ssenger compartment floor beams and connecting frames, cutting control cables and electrical wiring. And eventually loss of power. loss of control, midair disintegration.s Questions arose during the Aloha Airl ines accident investigation regarding certain tnforreat ion in the CAA report about cabin floor beam bending that suggested that the initial failure may have been in the upper lobe of the fuselage as opposed to the lower lobe as cited by the CAA. Testimony of Boeing and FAA experts at the Safety Board's publ +c hearing (See appendix A.) on the Aloha Airlines accident revealed that the evidence cited in the Ck~ report was consistent with an initial failure in the lower lobe of the FEAT airplane. A review of N73711's d+screpancy logbook, the f1 ight attendant cabin log. the line maintenance activity log. and the dispatch logs for the day of the accident revealed no significant entries prior to the accident.
LdP

Joint Design and Bonding History

The B-737 fuselage is divided into four sections with sections 41, 43~ and 46 comprising the majority of the pressure vessel. (See figUT4! 1.) These sections, along with section 48, are butt joined at circumfereht1al fraPes to fo~ the entire fuselage. Sect!on 43 forms the forward cabin area from 8S 360 to BS 540, where the area of skin separation occurred. The sections are constructed of circumftrential frames and longitudinal stringers that are covered by formed alUliinum skin panels that are riveted to the
T.;rcr.ft ac:efd.,u In ... sti,.tion •• ~ort (translation). 'ar fran.port. LTD., lo_in, 737-200. 1-2603. A"gust 22. 198'_ Chit .d.i~Jstratfoft •• i"fltry of Co•• un'cation •• Taipei. Tatwan. C~in •• 8Th• WOrdift. of t~i. ~robable tran.lated copy of the off$cS.t Eastern Atl" Au·on."tlea lepublic of

t~.

cau.e ~•• been excerpted •• erbatl., fro. after.ft Accfdent lnvestigatian Report •

14

underlying structure. Each skin panel in the upper lobe of section 43 is the length of the entire section -- about 18 f~et. Adjacent skin panels are joined longitudinally by overlapping the edge of the upper panel about 3 inches over the edge of the lower panel. The overlap (jOint) area is fastened with three rows of rivets and a bonding process. The center row of rivets secures the lap joint to a stringer underneath the skin, which. in turn, is attached to the circumferential frames by riveted clips. Below the window belt and in the lower lobe. the skin is connected to the frames between the stringers using riveted l-shaped brack.ets (shear ties). In section 43. the skin panel lap joints exist at S-4l and S-4R. S-lOl and S-lOR, and S·14l and S-14R in the upper lobe and at S-I9L and S-19R and S-26L and S-26R ir. the lower lobe. The upper lobe sk.in panels in section 43 are fabricated frc'!'i two complete preformed sheets of O.036-inch thick aluminum that are joined tJgether using a nhot~ bonding process. An acid etch is used to prepare the surfaces of the sheets bef~re bonding. Since the epoxy hot bonding material is nonreactive at room temperature, the bond is cured at 250°F at 45 psi (hot-bond precess). The inner sheet ;s then masked and the panel is mill ed chemically leaving the waffle" doublers that provides circumferential tear straps at IO-inch intervals and a lcngitudinal double thickness at each stringer location.
K

On the early model airplanes (through production line number 291) • . the doubler sheet was milled away chemically at the lap jOint locations; for production line number 292 and the subsequent numbers, the doubler sheet was retained on the outer panel of each lap joint to provide an extra 0.036 in:h . of material thickness in the joint. (See figures 3, 4a and 4b.) Additionally. for production line number 465 and the subsequent nambers, an improved bond surface pretreat process using a phos~horic acid anodize was employed.

For 8-731 production line numbers 1 through 291. the fuselage skin lap jOints were ~cold· bonded. A cold-bonded process used an epoxy impregnated woven ·scrim- cloth to join the longitudinal edges of the single thickness O.036-inch skir panels together. In addition. the joint was mechanically assembled with three rows of countersunk rivets. The metal surfaces to be bonded were etched to ensure cleanliness ar.d to prepare a suitable bonding surface. Since the epoxy ·cold- bond material was reactive at room temperature~ it was stored in rulls at dry ice temperature until shortiy before its use. It was then allowed to war'll to room te.perature before installation. This bond cured it roo. temperature after assembly.
The cold bonding process was intended to provide structural efficiency and Nnufacturing cost advantages plus overall airplane weight reduction over traditionally riveted thick skin panels. Fusel.ge hoop loads (CircUMferential pressurization loads) were intended to be transferred through the bonded jOint. rather than through the rivets. allowing the use of lighter. thinner fuseJige skin paneh with no degradation in fatigue life.

"

15
UNE NUMBER 1-291

HOT80NDEO

TCARSTFIAP

r---r
{,

':t

;.,

HOT BOND

CURED AOHEsr IE COlD BOND

CLOTH

..!-- SCRIJii

1
i

I
--,~~~_

HOT BONO£O

TEAR STRAP
NOTE: SINGLE THICJ(HESS AT BOTTOM OF PANEL

•
NOT£: ~

LINE NUMBER 292 II AFTER

CHANGE. DOU8lE THKXNESS

AT BOTTOM Of: PANEL

Figure 3.·-8-737 Lap splice conf1gvr~tion

16

UNE No. 1.. 91 2

UNE No. 292-AND AFTER

•
NO'rC: .. ..,,101C1U ....
1JCIICNDI CI1ENICW 0.. ..

1MaINE ••

1IURaID1'01IC.W1FM.

Ftgure 4a.--Lap joint section .. Ween tearstraps

17

UNE No. 1·291

UNE No. 292·AND AFTER

MIlAM
.....

I I I I I

\,--

1&-

--,

...

wtR!D

::-...... , I

"
I I

FflU" O.··Lap Joint sectton at t •• rstraps

18

Laboratory hut"l(J full Company and the results

wcoupon" tests9 of the bonded joints~ as well as the -Quonset scale fuselage section fatigue test were performed by the Boeing were used to assess cold bond durability. According tc Boeing, indicated that certification requirements were met.
8

The early service history of production 8-737 airplanes with cold-bonded lap joints (plus 8 727 and 8-747 airplanes with the same construction technique) revealed that difficulties were encountered with this bonding process. It was found that the cleaning and etching process used on the skin panels had not provided a consistent quality thin surface oxide to be used as a bonding surface. The service history compiled by Boeing has shown that bond quality can also be degraded if condensation is not removed from the scriin:cloth before inst~llation or if the scrim cloth sits at room temperature too long causing it to cure prematur~ly. According to Boeing engineers these production process difficulties resulted in the random appearance of bonds
t

moisture could enter the joint in the areas of disbend, and corrosion could occur. The moisture and corrosion in some cases contributed to further disbanding of the jOint because of the accumulation of oxides, water wicking in the joints, and the freeze-thaw cycles. The cold-bond lap joint production process on the B-737 was d;scontinued by the manufacturer in 1972. (See figure 5.) A redesigned smooth. close-fitting, Wfay" surface sealed lap joint with increased jOint thickness was intreouced with 8-737 production line number 292. This is a riveted Joint with chromated polysulfide sealing compound, but it contains no bonding. Production of 8-727 line number 850 ' and subsequent numbers and B-747 line number 201 and subsequent numbers also included fay surface sealed lap joints. According to Boeing engineers, when disband occurs in the bonded lap joint, as designed for the 8-137, the hoop load transfer through the joint is borne by the three rows of countersunk rivets that mechanically fasten the skin panal s together. Because of the single thickness skin surface that was facilitated by the bonded construction, the cour.tersink for the flush rivet heads extended through the entire thickness of the outer O.036-inch sheet. A knife edge was created at t~e bottom of the hole which concentrated stresses. These stresses were cyclic with pressurization loads, and fatigue cracking ultimately occurred at the site. In a cylindrical fuselage like the 8-737, the circumferential pressurization stresses are twice as large as the longitudinal stresses. As fatigue effects take place, cracks propagate longitudinally, perpendicular to the dominant pressurization (hoop) loads. In the 8-737, fatigue cracking
'·Couponconfiguration. describes s.all sections of akin ai.ul.tfng the joint

with low environmental durability, with susceptiblity t~ corrosion, and with some areas of the hp joints that d'ic! not bond at Once;n service,

an.

fO·Ouonset hut" refers to • full scale 1/2 section of fusela,. containing botb the upper and lower lobe.

1117 ,.

,.

1m 117111721m 11741m

1m

I -~---------I!-

s.

IN"JOIHTS RU.£f

SEALED

I lINE
4415 7178

• FWan'

PROBLEM fi90RTS __

&AI" JOINT COMOSIOff ... , .mJ

'ltllEPORT ... ! nn4

(WINDOW BELT1"0

l£M STIW' DIS8OHO

s-m -, MUlTft£ 11 ftEPOr!TS CRACI(S _'\_ .L r.l

3113101M M7

iN' JOINT

.~1Ot7 COlDIOHO SEALING CI"NITIAllY

EFF£CT1YE)

....
Figure .S.--Product1on, first probl_ reports, and
Service Bulletins/ADs for 737 bonded body skin lap jOints and tear straps

,,-

20 initially is expected to occur in the outer layer of skin along the lap joint because the outer layer contains a knife edge at each of the countersunk rivet hol es. Furthermore, the fatigue cracking primat';ly is found in the upper row of the outer skin panel lap joint rivet holes because this area carries the greatest stress. For the underlying skin of the lap joint, the area of greatest stress ; s through the lower row of 1ap joint rivet holes. However, since the rivet holes in this skin panel are not countersunk, fatigue cracking is not as likely to initiate at this location. Random cracking at lap joints (See Section 1.17.4~ Service Difficulty Report Information) on individual B·737 airplanes has occurred over time, related to the original quality of the joint bond and the environment in which the airplane operated. The rate of crack propagation has been d~pendent on, among other things, the degree of disbond at tnt! given location within the joint and the accumulation by the airplane of equivalent full pressurization cycles. During the service history of the B-737, Boeing issued several service bulletins (SBs) pertaining to corrosion detection and repair on fuse14ge skin panels, lap joint corrosion, disband and repair, and lap joint fatigue cracking inspection. The earliest of these was S8 737·53-1017 dated May 13. 1970, "Seal ing Of Cold Bonded Structure For Corrosion Protection. Two years later, the information was moved to the structurat R~pair Manual and the S8 was delated on July 20, 1972. As a follow-up, S8 737-53-1039 was issued on July 19. 1972, and initially addressed the area of lap joint . corrosion and repair on the first 291 airplanes produced. TOis 5B received a minor revision in October 1972. A revision/reissue in February 1974 reported lap joint disbond and corrosion on 30 airplanes and stated "in most ' instances these areas could be positively identified only after corrosion caused exterior skin bulges, cracks or missing fastener heads,· and "prolonged operation with large areas of delamination (1isbo~ding) will eventually result in fatigue cracking.~ The S8 pr~~ram outlined "the minimum requirements for maintaining the structural integrity of the lap jOints." Corrosion and fatigue inspection details and intervals and repair instructions were presented. Operator compliance was not made mandatory by the FAA.
It

'

On August 20, 1987, the subject 58 was elevated to "Alertlt status with Revision 3. The following was reason for the upgraded status: S inee the release .of Rev; sion 2 an operator has reported multiple fatigue cracks on three airplanes which have accumulated 40,400/42,800 flight hours and 44,700/49,900 flight cycles. Cracks were located in the upper skin at stringer four (5·4), left and right, S-10 right and S·14 right, between Body Stations 360 and 907. Therefore, Revision 3 was issued to up-grade this service bulletin to an "ALERT" status and to revise the repeat inspection thresholds for detecting fatigue cracking of the outer skin panel at the lap jOint upper row of fasteners.

21

Part I of sa 1039 Revision 3 dealt with "Corrosion IMcpection" and Part II addressed "Fatigue Damage and Repair." The subject areas were lap jOints at S-4, S-10, 5-14, S-19, 5-20, and S-24, Part III covered -Tear Strap

Inspection and Repair in the same structural areas as Part II.

K

The FAA issued an Airworthiness Directive (AD) 81-21-08 effective Novem~er 2. 1987, which stated in part: To prevent rapid depressurization as a result of fatlu~e of certain fuselage lap splices, accomplish the following: ...• (instructions followed) The AD made the inspection for fatigue crackil'lg referenced in S8 737-53AI039 Revision 3 mandatory for S~4l and R (note only S-4l and R) on product ion 1ine numbers 1 through 291, before the accumulation of 30,000 landings or within the next 250 landings after the effective date of the AD. Repairs for cracks found were to be accomplished in accordance with instructions contained in the referenced Boeing SB. (The AD and SB revisions 2 and 3 with nondestructive testing (NOT) instructions. are included as appendix C.)
t

An additional S8 737-53-1016 dated October 30. 1986, deals with skin bonding problems. (A summary of S8 737-53·1076 is included as appendix D.)

Boeing issued revisian 4 to 5B 737-53AI039 dated April 14. 1988, to permit an interim repair when cracks were detected and time was not available for complete restoration per the previous tnstruct.tons , This information was not relevant to the accident.
1.6.3 1.6.3.1

Aloha Maintenance tUstory Maintenance Program

Airpla~es operated by Aloha Airlines are maintained under an FAA-approved Continuous Airworthiness Maintenance Program as required by 14 CFR Part 121, Subpart l. The maintenance. based on guidance provided in the Boeing Maintenance P'I anning Document (Mpr;; (Document number 06-17594), recommended that aircraft maintenance inspect.Jns be divided into four series of checks with specific recurring frequency. The checks are referred to as follows: A B C D Check--Primary inspection to disclose general condition Check--Interrnediate check to determine general condition Check--System and component check, airworthiness evaluation Check--Structural inspection, determine airworthiness

-~

22
A Boeing study of early MPO documents revealed over-lap between C and 0 check items. Revision A of the MPOin 1974 redistributed tas~ items to other appropriate check intervals and the 0 check terminology was eliminated. However, no maintenance items were deleted and many airlines, ~1l.'lcludin9 Aloha, continued with the original terminology. Table I.·-Frequency of Inspection. (by n 19ht hours)
~

Boeing R~~ommendat;gn A B C

Industry

'12~71

Average

Schedy'~

Aloha

3,000 20,000

125 750

150 650 3,000 21,000

3,000 15,000

150

175

The Aloha Airlines work schedule for 0 checks initially was prepared h) 1972. The tasks from the Boeing MPO were organized into 52 increments (blocks) to be accomplished during the 0 check interval. Thee check tasks were organized into four increments and integrated with the B check schedule of work. B, C, and D checks were actually combined and accomplished in overnight segments. Aloha Air1 ines was participating in the Supplemental Structural Inspection Program (SSIPJ for large Transport Airplanes in accordance with FAAAdvisory Circular (Ae) 91-56 dated May 6, 1981. The SSIP is a continuous structural inspection to identify crack.s, corrosion, and other damage. While the program is not a substitute for the operator's !xi st i n9 FAA-approved structural inspection program, the SSIP clnd. the Supplemental Structural Inspection Document (5S[O)1' provide the operator with procedures to evaluate and supplement their existing program. The5SID provides for the inspection of Significant Structural Items (551) that have damage or fatigue characteristics that could affect the a'irpl ane' s structural integrity. Should cracking occur, the examination of SSIs allows operators to detect fatigue damage before the airplane's residual strength falls below the regulatory fail-safe requirements. (See 1.17.5 Supplamental Structural Inspection Program.) Though not related to, airplane fuselage skin in section 43, the review of the maintenance records found several SSID items for which no maintenance entry could be found. These SSID items were F-20, F-22B, F-24B, F-29A, and F-29b which pertained to the inspection of bulk.heads and door or hatch frames. Aloha Airlines personnel reported that these inspections had been incorporated into its letter check maintenance program. However, the
"IO.lno, with a.sistance frOIl the operatfng airlines, developed and pre •• nted to the FAA, proor ••• t~ extehd the operating life of older airplane. and to ensure tbe contfnu.d s.fe operation of those .irplanes. The FAA I.sued AO 84-2"06 ~ffectiye lov •• bet "84 to place the progra. in eff~ct fDr tt'''' .-737.

'23

Component Historical Record card for these inspections showed no such maintenance entries to indicate that the inspections had actually been accomplished. Maintenance Records Review To review N73711/s most recent complete cycle of A, B, C, and 0 checks, the Safety Board examined airplane records from May 15, 1980, to April 28, 1988. Aloha Airlines aircraft utilization was such that 8 years of flight activity was necessary to accumulate the 15~OOO hours which constitute the 0 check inspection interval. There are eight structural inspection blocks (portions of the complete 0 check) that require the removal of airplane interior components. These inspection blocks were proposed by the airline and approved by the local FAA principal maintenance inspector (PMI) to be accomplished sequentially, one block at a time. A one-time heavy maintenance hanger visit for a 0 check was not scheduled. A complete
interior removal at anyone time was not required nor was it accomplished.
The maintenance records review tndtcated that the prev ious cycles of A, B, C, and 0 checks were recorded as accomplished ~ithin the prescribed intervals. The most recent scheduled maintenance checks were: A--April 25, 1988; B--March 31,1988; C-4--March 31,1988; D (bloct. 5)--June 22,1987, (This block called for inspection of fuselage skin and framing around windshields and windows.); and 0 (block 8)--February 20, 1981, (This block called for inspection of fuselage skin and stringer splices at BS 320 and a general inspection of the fuselage at BS 400 and BS 520 areas) .

1.6.3.2

The 0 check structural inspection included an FAA-approved i/4 sampling program. This meant that certain blocks of the D check wer-e accomplished on 1/4 of the airplanes in the Aloha IO-airplane fleet at the normal 15,OOO-hour interval, and if no adverse findings were encountered, another 1/4 of the fleet was inspected at 30,000 hours. Aga in, witn no adverse findings, another 1/4 was t~ be inspected at 45,000 hours, etc. lhe Boeing MPO states, Should an operator encounter an adverse finding, the following actions are recommended: (I) Inspect remaining aircraft in his fleet at the earliest opportunity. (2) Eval~ate findings from these inspections together with data from Boeing on the inspection time or area, (3) Determine if a change in frequency of the time interval and/or the fraction needs to be accompl ished and then make the change in the program. There were no adverse findings recorded in any of the records reviewed; therefore, there were no changes in the frequency of inspection or the fraction related to the sampling program.

•
,

After the accident, the Safety Board conducted visual inspections of the exterior of the airplanes in the Aloha Airlines 8-737 fleet. Considerable evidence of corrosion on the fuselage of the airplanes in the

------_

...

_-_ .._-

.--

..-.. ---~

24

fleet was seen. Swelling and hulging of the skin (pillowing), dished fa~tener heads, pulled and popped rivets, and blistering, scaling, and flaking paint were present at many sites along the lap joints of almost every airplane. Aloha Airlines did not produce evidence that it had in place specific sevare operating environment corrosion detection and corrosion control programs employing the techniques outlined in the Boeing Commercial Jet Corrosion Prevention Manual (Boeing Document 06-41910). Program requirements in the manual include extensive application of water displacing corrosion inhibiting compounds, reapplication at fastener locations and panel edges of exterior fuselage ski n every 6 months and interna 1 treatment at 2-year intervals, wash; 09 the aircraft at IS-day tnterval s, plus regular buffing and brightening of the unpainted surfaces. Aloha Airlines maintenance 0 check instructions for structural inspection addressed corrosion with an introductory note. This notation defined the inspection as a rigorous visual examination for condition (damage, cracks, galling, scratches, wear, corrosion, rust, evidence of overheating, rubbing, or age) without further definition. Aloha Airlines inspectors and quality control personnel stated that the corrosion was corrected when detected during normal inspection and maintenance activities as part of their normal task card act tvi ty , The Safety Board subcategorized and evaluated a11 pressurization discrepancies recorded from 1980 to 1988 to determine adverse trends or significant anomalies. This maintenanc~ historical review produced no e~idence of prior structural overstress incidents for N73711 as a result of pressurization or other ~alfunction. 1.6.3.3 Servfce Bulletins Boeing periodically issued information via SBs to inform operators of reported or anticipated difficulties with various airplane m~dels. The following communications were relevant to the B-737 fuselage structure, including section 43: o o o o o Structural Item lnterim Advisories (SIIA) Service Bulletins (S8) Service letters (SL) In~Service Activity Reports (ISAR) Significant Service Items (SSI)

Nine SBs provided guidance for maintenance or information otherwise applicable to section 43. Of these nine SBs, entries referring to the foilowing five SBs were found in the Aloha Airlines fleet maintenance records: .

SB 737-53AI042 Lower lobe Skins S8 737-53AI064 Frames Stations 351 and 360

58 737-53AI027 Cargo Compartment Body Frames S8 737-53AI039 Skin lap Joint Inspection

S8 737-53-1017

Sealing of Cold Bonded Splices

24 fleet was seen. Swell ing and bulging of the skin (pillowing). dished fastener heads, pulled and popped rivets, and blistering, scaling, and flaking paint were present at many sites along the lap joints of almost every airplane. Aloha Airlines did not produce evidence that it had in place specific severe operating environment corrosion detection and corrosion control programs employing the techniques outlined in the Boeing Commercial Jet Corrosion Prevention Manual (Boeing Document 06-41910). Program requirements in the manual include extensive application of water displacing corrosion inhibiting compounds, reapplication at fastener locations and panel edges of exterior fuselage skin every 6 months and internal treatment at 2-year interval s , washing the aircraft at IS-day intervals, plus regular buffing and brightening of the unpainted surfaces. Aloha Airlines ma intenance 0 check instruct i ens for structural inspection addressed corrosion with an introductory note. This notation defined the inspection as a rigorous visual examination for condition (damaget cracks, galling, scratches, wear, corrosion, rust, evidence of overheating, rubbing, or age) without further definiti on. Aloha Ai r l tnes inspectors and quality control personnel stated that the corrosion was corrected when detected during normal inspection and maintenance activities as part of their normal task card
activity. Board subcategorized and evaluated all pres.sur+zat ton recorded from 1980 to 1988 to determine adverse trends or Significant anomalies. This maintenance historical review produced no evidence of prior structural overstress incidents for N73711 as a result of pressurization or other malfunction.
di screpancies

The Safety

1.6.3.3

Service Bulletins

Boeing periodically issued information via SBs to inform operators of reported or anticipated difficulties with various airplane mDdels. The following communications were relevant to the 8-737 fuselage structure, including section 43: o o o o o Structural Item interim Advisories (SIIA) Service Bulletins (S8) Service letters (Sl) In-Service Activity Reports (ISAR) Significant Service Items (S5I)

Nine SBs provided guidance for maintenance or information otherwise applicable to section 43. Of these nine SBs, entries referring to the fOllowing five SBs were found in the Aloha Airlines fleet maintenance records: .
S8 737~53-1017 Sealing of Cold Bonded Splices 58 737·53AI027 Cargo Compartment Body Frames 58 737-53AI039 Skin lap Joint Inspection S9 737-53AI042 lower lobe Skins S8 737-53AI064 Frames Stations 351 and 360

41

25

Due to the method of entering the SBs in the Aloha Airlines maintenance records, the recurring nature of inspections could not be detennined. Also, entries for the following four SBs were not located in the records:
S8

737-53-1076 Repair S8
737-53-1078

Fu~elage-Bonded Skin Panel Iospection and Fuselage Window

Belt

Inspection and Repair

Skin

Panel

S8

737-53-1085 fuselage Stringer to Frame Tie Clips Inspect;or. and Replacement Fuselage Skin Crack At Strir.ger 17 Inspection and Preventive Modifications

SB 737-53-1089

Aloha Airlines personnel stated that the information contained in these particular SBs had been incorporated into Aloha Airlines letter theck inspection system; however, specific documentation of this fact was not produced.
\

1.6.3.4

FAAAirworthiness

Directive

(AD) Compliance AD 87-21-08

Airworthiness Directive (AD) 87-21-08, which became effective on November 2,1987, was issued "to prevent r-apid depressurization as a result of failure ofeertain fuselage lap splices .•••R The AD required operatcrsto perform a "c'lose visual inspectionRn of S~4l and R. and if cracks were found, operator-swere required to perform an eddy current inspect ionllof the skin around the upper row of lap jOint rivets for the full length ofth~ panel. Compliance with the AD was required before the accumulation of 30,000 landings or within 250 landings after the effective date. whichever occurred later. The AD was based on Boeing Alert Service Bulletin (ASS) 731-53AI039, Revision 3~ dated August 20, 1987. The ASB required an inspection of the skin around the upper row of rivets along the lap joints at S-4. -10. -14. -19. -20. and -24 left and right. An FAA employee testified at the public hearing that the decision to limit the seopeof the INndatory inspection was based on analysis of statistical information available to thea and the recognition of the scope of work required. A review of the maintenance discrepancy logs found that two repairs to cracks on the S-4R lap jOint on N73711 were accomplished on Novellber12. 1987. The small separated section of upper fuselage recovered after the
12T".
·clos. S.fety to. ret lUIS vi.u.' f,.ap.ctlon.·

unable h.

to

toeete

aft tndv.try te.t

d.fintt~on

of

13,," etJdy cur,..nt

inspection

nondeat,.uc:ti"e,

(.Of)

•• tho.'"
object. A c.u ••• the

It"teh an fr:dlleed ttl.etrical eddy curtel\t fa ,.neratect tl\ the t.at •• terial deviation such as • cract or difference In akin thfctn~s. .dld., current to chan,. and altolf. the ano.a:y to be detected.

25 Due to the method of entering the S8$ in the Aloha Airlines maintenance records, the recurring r.ature of inspect ions could not be deteJ"'lllined. Also, entTies for the fo-:lowing four S8s we1"'e not located 1" the records:

SB 737-53-1076 fu~eldge-Bonded Skin P~l

Repair
S8

Inspection and Skin Panel

Inspecticn and Repair

737~S3-1078

Fuselage

Window

Belt

S8 137-53-1085 58 137-~·1089

lnspectior. and Replacement

~nd

Fus~lage Stringer

to Fraae lie Clips 17

Inspection

Preventive Modific~tions

Fusel~~

Skin Crack At Strir.gtr

these particular SBs had been incorpor~ted into Aloha Airlines letter check inspect ion sJstee; ",*"er.. specific docu.t\tat ion of this fact 8S ..,t
produced.

Aloha Airlines personnel stated that the infor.ation contained in

1.6.3.4

FAA Ai~rthiness

Directive (AD) eo.pliance AD 87-Z1-08

Atrworth~ness Oirective (AD) 87-21-08 .. "Meb bee... ffecttft Oft Icwtllber Z~ 1987,. lAS. is$.ued "to ,I'ftfttt rapid deoressurilI.UOft .$ a. result of faUa,.. of cmaiD fasel. lap spUces •. _... TINt ADrequired ooentc1"S to perfOftl a ·close vhual taspectiOft·n of S-'l and ft. aM if c1"acks ..,.. fOGtld~ opentors ~ f'e(\ViAd to per'fonJ aft ~ c.t'Teftt ias.pec.tiQftU of tM $kift arowad the IIPger of lip jot:tt rhttts for tM f.11 l~ of the ,...1. CoIIpl1ac. vUh t....AD.as requind before accu.alaUon of 3O~0D0 lat\diftlS or wttb"t '2SO 11ftd1* aft.,- tM .ff.ct1ft ot.,. wd~ oter,," 1~. 11Mt AD was INnd Oft 8oIi19 Alert Sefttc netta lASe) 731..Sll1039ft AevisiOft 3. claUd Augut 20" JI81. n. AS8 im .,. fupftU_ of ... ski. IrGaIId tile __ yo.,..,.. of .. 1..u .1001 t.he lap .lola\$. at $-4. -10. -I". -19. -20_ lid ·24 l"t ... right" ... FAA ..,10Jft bstift_ .t tM ... 1t~ hViag tht tk dec.'st_ to 11... t u. SCOPt of tM "-t0t'7 las.pect1oa was based on INl",U, of statist'cal i"f.-tioa •• to u.a ... ttMt

"*

t_

teCOgft1tloo

.,

UW HOpe of

A twl .. of logs f~ tNt two to CriCks Oft U. $..41 lip jot At Oft 111111 ...... Kc_lhMd on _tIIIIbtr II., J981~ n. _n '....,.'" s.ctt_ of .,... 1-..11 .. nco.,,"", .n.r tile

t_ ..,.,
t ...

,....'rv4.
u. ,.
't

,..,,"....,
tfl ••••

,n.,.
'~t

"'*,'"
.f •

-.,

••••

1".,,,,. "._'.' '

f ... '
I"'

1~,.tti.4~·

••• ~

It:

"f'''ft·.~
t •• ,)
, •• t tft••

1J.~ .~~
... ~.'"

e.~~.~f ~-t•• ftt.~

~ , ;o>-J •• ••

.. t.tr..~ * ~~.~ ~ •• ~ cy~r.~f~ •••• ~•• _*4 •• ~••• , ••• ~... '~

t., .•~... ~.;

fl ...

.,

It ..."..~ .-,

~.I!'t" ..... .. .ff'.*.~.~~

; ~.~••t~ .. t' •• t •• t
t ••

.tt~

t ...

_., •••

t~'~&~ ••• c •••••••• .c .

"'M'. •

.. t ...

,_

--

-~

26

a.ccident contained both of the repaired areas. The maintenance log for N7371l indicated that a visual inspection had been accomplished in a~cordance with AD 87-21-08; however, the record contained no evidence that the required eddy current inspection had been accomplished. An Aloha Airlines inspector testified at the public hearing that it was company practice to perform an eddy current confirmation inspection whenever a crack was detected visually. Both the Aloha Airlines directcr of quality control and the staff vice president for quality assaranc~ and engineering stated that a Nondestructive Testing Report (Form No. M-86) should be filled out by the inspector when any NOT inspection is performed. The form is then used by management for tracking purposes. A search of the records for N73711 failed to find a copy of an NOT inspection report of the S-4R lap joint. The inspector who performed the initial AD inspection on N73711 stated that he did not believe that document~ng the eddy current inspection was necessary or required. During the investigation, Aloha Airlines did not produce a written maintenance policy regarding the requirement for the entry of an eddy current inspection in the rna ntenance 1;)9. However, a broad i examinat ion of rna ntenance records revealed that other inspectors had made i such entries during this same tiIDeperiod. Two insp2ctors wo~king on separate shifts conducted the inspection required by AD 87-21-08 on the accident airpla~e. They followed guidance in the AD and the related SB (SB-737-53AIU39) which were taken to the work site. The first inspector started on November 12. 1987, and visually detected the cracks on S-4L. This inspector stated that after visually detecting the cracks, he performed an eddy current inspection of t~e lap joint upper rivet holes along the length of the panel (BS 360 to as 540) and found no additional cracks. After maintenance personnel accomplished two sheet metal repairs, the first inspector inspect~d the work and signed the log book. The second inspector stated that he performed a complete visual inspection of the airplane, including the area inspected by the first inspector and the two repaired areas, and he signed off the completion of the AD in the maintenance log on Kovember 14, 198/. The related inspections on the lap joints at S-lO~ -14, -19, -20, and -24, which were recommended by S8 737-53AI039 but not by the AD, were not accomplished. At the time of the AD inspection and repair, N73711 had accumulated 87,056 cycles. The accident occurred at 89,680 cycles.

1.7

Meteorological Information

The accident occurred in day visual meteorological conditions. There was no significant adverse weather experienced.

1.8

Aids to Navigation
Not relevant to this accident.

27

1.1

Communications

There were radio communications d~ffieulties between flitht 243 aot ATC shortly after the explosive decompression. At 280 to Z90 knots lAS and with i part of the forwarti cabin structure and the eoc"-.pit door" Misstiig, b .. _ noise levels impeded air/ground communications briefly. There weTe no other communication ano~lies. 1.10

Aerodra.e Inforaation

After the explosive decompression, the airplane proceecfecfto the nearest suitable landing field, Kahului Airport, a 14 CFR Part 139 certificated Index 0 airport on the island of t1aui Hawaii. The only instrument runwaYt 02/20, is 6,995 feet long, 150 feet ~ide~ and constructed of asphalt with a grooved surface.
y

1.11

~light Recorders

The airplane was equipped with a Fairchild model 5424 foil type :ana10g FORt SIN 7274, and a Coll ins model 642C-l cockpit voice recorder (C.VR), SIN 54. After the accident, the recorders were ~ved from the airplane l~d sent to the Safety Board's flight Recorder liboratory ift Washington, D.C. for examination and readout of pertinent data. (See append ix E.)

Examination of the FOR recorded traces indicated that the flight was normal fr~~ liftoff to the accident. The airspeed trace abruptly ceased at the time of the accident and dropped to a position below zero KIAS. The other recorder parameters appeared to operate norma 11y. Peak vertical acceleration (G) excursions recorded as a resuit of ·the icciden-t were ·0.'8 and +2.95. These peak values were not sustained. The CVR revealed norma" cOlllnunicationsbefore the decOIIPression. Following the decompression, 10Qd wind noise from the openin9 in the fuselage ,revented normal cockpit conversations. Hand signals were used to communicate. When the a~rspeed and related wind no1s~ had been reduced to a level where conversations were intelligiblew the fiightcrew discontinued using the oxygen masks. Cockpit ccnversations then continued to be recorded in the normal manner. 1.12

Wreckage and I.pact Infor.atton

The extensive air and surface search of the ocean failed to locate the portions of the airplane lost during the explosive decompression. 1.13 Medical and Pathological Information The flight attendant who was ejected from the fuselage was not found and she is assumed to have been fatally injured in the accident. Two passengers who wers seated in the first class cabin in seats 2A and 2C were struck by debris ifid wiring which resulted in .ultiple

28

Passengers including cerebral concussions and multiple lacerations to their heads and faces. Passengers seated in 48, C) 0, and E (center and aisle seats) sustained multiple lacerations and were treated and released on the day of the accident.

seated in seats 4A and 4F (window s~ats} sustained serious injuries

lacerations

and electrical

shock burns to the face and hands.

concussions

Passengers seated in rows 5, 6, and 7 also sustained cerebral and multiple lacerations. An 84-year-old female passenger seated in SA was the most seriously injured with a skull frac~ure, lacerations and a skeletal system fracture. The passenger seated in 6A sustained a broken right a~f multiple facial lacerations, and blQod effusion in both ears.

The majority of the passengers ~eated in rows 8 through 21 received including lacerations, abrasions, and barotrauma. They were treated and released on the day' of the accident. Twenty-five passengers reported no injuries and continued to thp.ir destinations that same eve~ing. There were no reported injuries as a result of using the emergency evacuation slides.

a1nor injuries

1.14

Fire There was no fi~e.

1.15

SUrvival Aspects •

This was a survivable accident; the fatality was the result of the explosive nature of the decompression. The flight attendant was swept violently frOil the airplane and passed through an opening of jaggeJ metal. There were blood stains on seat cushions at seat SA on the left side of cabin near as SOOand on the exterior left side of the fuselage where the flight attendant was standing when the decompression occurred. Passeng&rs who observed her during the explosive decompression stated that they saw the flight attendant pu1,ed upward and toward the left side of the cabin at seat row 5.

1.15.1

SUppla.ental

OXygenSystems and the cockpit observer seat occupant used the

airplane-installed crew oxygen system. Postaccident inspection showed that botb the crew and the passenger oxygen bottles that were located in the forttard cargo cOIIPartllent had zero quantity and pressure. The passsnger

The f119htcrew

omen

structural separation, and thus, first-class and coach cabins.

J.15..2
At 1430, the

distribution

.. nifolds

were part there

of the material lost during was no supply of oxygen to

the the

Airlines t·731 was diverting to

A Coa$t

search the area for debris

fAA notified the U.S. Coast Guard that an Aloha Maui airport dee to an ftinflight explosion.Guard helicopter, airborne on a train~ng mission, was assigned to

and the flight

attendant.

The Coast Guard cutter

29

tAPE CORWIN was also directed into the search area as was a Marine Corps helicopter. A full search effort by ships, helicopters, and fixed~wing aircraft con~inued fo~ 3 days without success. 1..5.3 1 Rescue and Firefight;ng Rgsponse responded with five emergency had evacuated the airplane via personnel entered the airplane occupants were removed from the The Mau; Airport fire department "Iehicles. After the ambulatory passengers slides and the aft airstair, fire depart~ent and assisted the injured still on board. All airplane in 25 ~inutes. 1.15.4 Ambulance Response

The flightcrew initially communicated the nature of the emergency as a "rapid decompress ton." The full nature of the structural damage was not verbalized. ATC notified rescue and firefighting personnel, but did not immediately call for ambulance assistance. A subsequent call from the flightcrewat 1353, "We'll need assistance for the passengers when we land," was confirmed by ATC personne1. Pol ice dispatc.her records indicated the "Medic Iff ambulance was notified at 1358, about the time of touchdown. A reason for the notification delay was not determined. The first ambulance arrived at the scene at 1405 and radtoed. for ass+stance. Other ambulance vehicles arrived at 1411. 1.16 1.16 .. 1 Tests and Research Pressurization System

All of the pressurization system wiring from the selector panel to the pressure controller to the outflow valve was examined. No dis~repancies were found. Additionally, a visual examination of the components including the outflow valve, both relief valves, the controller, and the selector panel did not reveal any discrepancies. These compcnents were removed from the airplane after the accident and subjected t~ standard Itceptar.ce test procedures for new units. There were no significant anomalies discovered. 1.16.2
Eddy Current and Visual Inspection

An Aloha Airlines inspector under supervision of t~e Safety Board conducted postaccident eddy current inspections on selected portions of the remaining fuselage lap jOints to determine the extent of fatigue cracking of the skin along the top row of rivet:,;(the area of highest stress). The inspected areas included the left and right lap joints at 5·4, -10, and -14 from BS 540 to 8S 1016.
Initially, the skin around 53 rivets exhibfted crack fndications along S-4L and S-4R, some visu~11y detectable by paint cracks. To make the rivet heads more discernible, the eatnt was sanded off and the skin was reinspected. Twenty-eight of the original 53 indications were confirmed cracks. Stripping of the paint layers was not attempted. (It is not normal

30

Aloha Airlines or industry practice to remove paint by sanding.} Two samples of the lap joint were cut frfJlll between BS 727 and 747 and betweer'l 847 S-4l BS and 867 for further examlnation. The eddy current inspection along 5-10 and 5-14 revealed 17 cracks along S-lOl and 2 cracks along S-14R. There were no cracks ale>ng S-IOR or S-14L. No atten!pt was made to strip the paint layers. (Appendix f provides details of these inspections.) There were 25 locations where previous fuselage skin, repairs or rework had been perfnrmed. r.ost of these areas consisted of external doubler patches at various stringer and frame locations. In several areas~ countersunk rivets haL been replaced with universal buttonhead rivets in lap jOints, mostly in the lower lobe. (Appendix G provides a description of the repairs or reworked areas anrl their locations.)
1.16.3

Materials laboratory Examination

Selacted pieces of the fuselage skin and associated structure~ were returned to the Safety Board's Materials laboratory fo~ analysis, These pieces included lap jOint samples (S-4R~ S-4l and S-lOl) and a section of a circumferential butt Joint strap. 360 and BS 420 (found wedged The patches were removed to examine the holes for evidence of cracks. ~here was extensive fatigue cracking in the upper row rivet holes both under anJ between the patches. The examination found one of the longest crac~s on the airplane, 0.27 inch~ in this piece. This stringer section (S-4R) contained three areas where the tear straps are riveted above the primar'y lap joint, There was extensive fatigue cracking present in all three locations. Also, the entire cold-bonded lap joint had become di~bonded. There was light to moderate corrosion with severe corrosion (unrepairable deplntion of metal) in some areas. Nearly all of the hot-bonded te~r straps were disbonded in the vicinity of the lap jOint.
in the risht wing area), contained two external doubler patches.

The lap jOint sample, S-4R between

as

The lap Joint s~ples, S-4l from BS 727 to BS 747 and from as 847 to BS 867, each contained 18 columns Cr lap joint rivets. The laboratory eXUlination revealed fatigue crac~ing in the skin "ldjacent to nearly every hole fn the upper rivet row with the larger crack lengths located in the .id-bay areas (half way between two adja~ent circumferential tear straps). A c_arison of the final results of the postaccident on-scene eddy current inspection condueted by Aloha Airlines technicians and the Safety 80ud laberatory findings revealed that the on-scene eddy current inspection only successfully identified cracks larger than 0.08 inch. The laborat~ry eXMinltton found fhe cracks that aaeasured 0.08 inch (+/- .005). The postlccident inspectfon had fdentified only on~ of these five cracks. This crack-length inspection threshold of 0.08 inch varies frOll the Boeing NOT Manull which states, -This inspection can find cracks 0.040 or longer beneath the countersunk fastener heads •.•••

31

moderate corrosion was present on the previously bonded surfac~s.

The lap jOint piece, S-4L, from BS 519 to BS 536, exhibited fatigue crack.ing from 16 consecutive rivet holes along the upper row of lap jOint rivets. The largest single fatigue crack in one direction measured 0.18 inch from the knife edge of the countersink. (See figure 6.) The longest total combined crack length in both directions across a rivet hole (end to end of the crack including the hole) measured 0.53 inch. Both the cold-bonded lap joint and the hot-bonded tear straps in this area had disbonded. light to At the request of the Safety Board, Boeing performed a striation count on several of the larger fat;gue cracks from the skin along S-4R lnd S·IOL to determine age and crack. propagation rate. Although data could not be obtained from all the cracks examined, Table 2 provides the estimated number of cycles of crack growth found on the seven crack samples that provided suitable data. Table 2.--Striation counts on selected cracks from the lap joints along S-4R and S-IOL
Specimen
S-4R S-4R

location

Estimated number of cycles {+/-20%1

28.670 37.148 28.656 26~449 2.,,056 23,628 36.379

Crack length in inches

0.105 0.142
0.154 0.145 0.130

S-4R

It
8Xa.nned

S-4R

S-4R S-ICl S-IOl

0.110 0.161 strap section. froll BS 360 at S-7R fatigue cracks frona both sides of a line. The fatigue regions extended

An examin.tion of the butt reye~led circtlll'ferentially propagating rivet hole just forward of the. joint 0.09 inch abov~ the rivet hole and 0.03 The separated £f'is
fOT

inch be10w tbe rivet hole.

of the No. 1 engine control cables were also condition &nd failure .ade. The separation areas of the lb. 1 eftgtM control cules were cleaMd andex_1ned. Each bnalt exhibited CGrJ"Osion; only • few o( the 1n\1ividVll wires Wl!re rel.Uvely uuffectee. ~,. of the strands exMbite-:t c.orrosion daaage through .ost of the wire eli 1IIft..- • 1..11
Additional

InfONlt.ton

1.17.. 1

sener_l

lnspect.tOft of 0tMr Alotaa. AirUMS .t..,llMS

n. safetJ Board Y"QvilWd 2-,.11" &Jint ... al't\:t records of thI'M other hip-cycl@ t-n1s ~at .._., by Alol:a AirHnts--fl73712,. 1111713,. _, mn1. All of t... rwqvind A. B. c. .ndO (i'M!c1ts Mel betft signed off at tM _ropriat. i..I'Nals. t l~ $.!o'tltlM?~tal s.trvctural ift$pt!CtiOftS ..... atcouM!d for Ifni. tM t.ceptton of SSID iteas pertlining to bulu.uds and

~---

32

UPPER SKIN

CArnCAl· RIVET
ROW

o o

o oo oo

.~
0 0
LOWERSION

Figure 6a.--Typical lap Joint

I
CRACK ORIGIN

KNIFE EDGE

TYPICAL CRACK
PROPAGATION

Figure 6b. --Cross sect i on of a ri vet and sk:in

33

door or hatch frames whi en a1so ~re not addressed i" the CQIIIpOneIft Historical Record cards of the accidefit airplane. The same SBs that had been applicable to the maintenance of N73111 were applied to the three airplaR@$_ In accordance with AD 87-21-08,. N73712 had be~ inspected Oft November S, 1987. At the time of the inspection~ the airplane bad accumulated 32,642 hours and 87,551 cycles. No defects wet~ reported during that inspection. On April 9~ 1988, with an accmnulated 33,676 hours and 90,051 cycles, the airplane was hangared for heavy maintenance. It was the highest cycle 8-737 in the world fleet. Following the N73711 accident. the N73712 airplane received a thorough ~orrosion/fatig,e inspection and evaluation of the structure. It was determined that the a:rplane was beyond economical repair. It was dismantled on the site and sold for parts and scrap. On April 14 19B8. N7l7l3 had accumulated 32~C26 hours and 85t409 cycles and received tt s last A check. No discrepancies were noted during that inspection. The inspection required by AD 87-21-08 had been accomplished en December 15, 1987, at 83,488 cycles. In 1984, Aloha Airlines submitted a Service Difficulty Report (SDR) to report a 7 I/2-inch crack on this airplane. The crack was located along the top row of rivets along the lap joint at S-IOR. The discrepancy log e.,try referred to SB 737-53-1039 .. Following the N73711 accident~ H737!3 received a thorough corrosion/fatigue inspection and evaluation ~f tbe structure which indicated that this airplane was also beyond ecoMmical repair. The airplane also was dismantled on the site and sold for ~arts and scrap_
t

On April 27, 1988, N73111 bad accumulated 39,g86 hours and cycles and received its last A check, No discrepancies were noted during that 1;.spectio.~. The inspection required by AD 87-21-08 was accomplished on January 12, 193B, at 67,429 cycles. The MIB maintenGnce form stated that both a visual and ail eddy current iuspection had been accomplished. The entry showed that the fuselage CTOwn from S-4R to S-4l had been repaired at station BS 540. Also, c~rrosion of the forward sectl~n of the skin jOint on the left side from S-9l to S-4l resulted in repairs*
68,954

After the N73711 accident. W731!i r~i~ned parked for almost 6 months awaiting final disposition. It was then fl~ on a ferry permit to an independent aircraft overhaul facility for refurbishme~t. During initial inspertion after paint stripping, fatigue cracking was foood visually at multiple rivet locations on the S-14R lap joint at BS 300 and numerous tearstrap disbonds and skin cnrrosion sites we~~ apparent. All outstanding S8 actions and terminating (permanent) repairs for the ADs pertaining to the structure Wile accompHshed. The ·li~plaM! was out of service for about 1 year.

•
---------___"j

34 1.17.2 The 8-737 Fail-Safe Boeing designed Design 8-737 for an "econonic service

the 3-737 was certificated in 1967, federal Air Regulations required that the airplane's structure be capable of sustaining 80 percent of limit load1s with any complete or obvious partial failure of any single structural element. Howevel"', the 8-737 was designed to sustain full·limit load to account for dynamic effects. The fail-safe design criteria for the 8-737 established by the manufacturer required that the fuselage be abl e to withstand a 40-inch crack without suffering catastrophic failure. These criteria k'ere derived from an estimate of the maximumexternal damage expected to occur to the fuselage as a result of external damage that might occur from the penetration of projectiles produced by an uncontained engine failure. There was no consideration given to the joining of adjacent cracks which might dlii'l/e1op during extended service other than normal "st ate-ofvthe-er-t" fatigue evaluation. Boeing design included the placement of tear straps with la-inch spacing in the fuselage skin in both directions (longitudinal and circumferential) to redirect running cracks from external damage in a direction perpendicuiar to the crack. The fail-safe concept was based upon the theory that the redirection of a progressing crack would cause the fuselage skin to -flap" open, releasing internal pressure in a controlled .anner without adversely affecting the residual strength of the fuselage as a whole. Supporting the skin are circumferentially oriented frames spaced apart and longitudinally orie:-.ted stringers located 10 inches apart. Each area bounded between adjacent frames and stringers (20 inches by 10 inches) is considered a frame bay. The fail-safe design requirement was to allow for failure within two frame bays without compromising the structural integrity of the fuselage.
20 inches

20 years and to include 51.000 flight hours and 75,000 cycles.

the

At the time

life"14

of

blades located nearly side by side which we~e used to penetrate longitudinally the test fuselage section within two adjacent frame bays while it was unde!" full pressure. The guillotines produced an instantaneous 40-tnch separation in the fuselage skin with a break in the center tear strip. As anticipated by the deSign, the separation redirected itself eircUllferentially, produced a flap~ and resul ted in a controlled decOllPression. Similar results were obtained when the ~ui11otine test was

Boeing demonstrated the ability of the fuselage to fan safely two frame bays during certification of the airplane by -gu1l1otine· tests on a fuselage half section. The guillotine tests involved two IS-inch
within

oriented cfrcumferentially.

'4' •• 1".·. defInition for -.cono_lc ••rvic. lif.- r.qulr •• the .'rplane to .ctafn the.e Yalu •• (51.000 f~i8~t hour. and 15,000 cycl •• ) wfthout .trvetr.tr.' feU,,,. crackin, which Mould (au.. "Iltificant operator •• tnten.nee axpan •••

35

During the certification program for the 8-737, some of the knowledge gained on the 8-727 full-scale certification testing was used by Boeing to validate fatigue performance on the 8-737. Skin thickness of the 8-737 (0.036 inch) was siightly less than that of the 8-727 (0.040 inch). However, fatigue testing of a complete B~737 was not accomplished as it was on the 8-727. That is. the complete B-717 fuselage was cycled for 60,000 cycles (one economic design life goal) during certification whereas the 8-737 fuselage design concept was demonstrated by fatigue testing a. representative crown-to-keel half section of the fuselage. The t~st section for the 8-737, or "quonset hut," was cycled 150,000 times to full pressurization differential (two times the 75tOOO-cycle economic design life goal). No fatigue cracks devr-Ioped on the test section and no disbanding occurred. These test results were used to verify the B-737 fatigue life expectations. There was no consideration given in the fatigue evaluation to the possibility of disbonding or the effects of corrosion on the strength of the fuselage lap joints.

1.17.3

In-Service

~del

Fuselage Tests

In 1986, Boeing acquired a B-737 that had been involved in an in-service accident. At the SClfety Board public hearing, Boeing personnel stated that the airplane was purchased for two reaSbns: to conduct a thorough teardown of the airp1ane from nose to stern, from wing tip to wing tip, fuselage, wing, empennage; and a 1so to run some damage tolerance test in9 of the aft fuselage, since the aft body was in good co"dition, to verify some areas we wanted to understand further about pressure bulkheads. The fuselage, line No. 90~ was acquired with jl!st over 59,000 actual flight cycles. lap joint and tear strap bonds were inspected and found to be in good condition. Boeing then applied over 70,000 additional test cycles. The first skin cracks (seven) located around BS 780 were disGovered in August 1987 by NOT at 79,000 cycles. In September 1987, It 89,000 cycles, there were about 15 cracks' detected in a 20-1nch bay area around as 820. The cracks ranged from about 0.37 inch to 0.67 inch tip to tip. At this po tnt., Boeing engineers placed additional straps on the test article at 8S 760 and BS 820. At the Safety Board pub11c hearirg, Boeing indicated this step was taken to preserve the test article in th~ event of I catastrophic failure. Boeing further indicated that the add~ straps would not alter the results of the fatigue testing. When additional cycles were applied, individual cracks joined to fora a large crack that grew to about 32 inches at 100tOOO cycles. Testing continued to 100,673 cycles; when the crack reached almost 40 inches, the skin flapped and controlled pressure release occurred. During the latter portion of the testing, the structure and skin yielded (deformed), and the crack gap remained open with interior insulation material visible after each fs;l1 pressurization cycle.

36

1.17.4

Service Difficulty Report Infor.ation

The FAA SDR data base was queried by the Safety Beard after the accident for information pertaining to the 8-737 fuselage. From the beginning of the current data base (January 1983) until the date of the accident, 1,352 records were found. Of these, 198 were reports of fuselage skin cracks, and 10 of these reports were of cracking at or near lap joints. Six of the 10 reports involved lap joints in the upper lobe, while the remaining " reports indicated cracks from the lower lobe of the fuselage. (See Appendix Ii.) One report was submitted after the effective date of AD 8J-21-OB, November 2, 1987. All of the airplanes cited in the 10 reports were among the first 291 8-737 airplanes assembled by Boeing. There were 18 SORs on file pertaining to airplanes in the Aloha Airlines fleet. Three reports were on lap joint cracks/corrosion ~reviously cited, and two reports were about upper lobe skin cracking whp.re lap joint involvement could not be established from the information given. Three of the =-eports pertained to lower lobe skin corrosion; an ~dditional three reports cited corrosion at cargo door frames and the nose gear wheel well structure. The remaining seven reports invol~ed cra~ks in fuselage structure other than skin or lap jOints. 1.17.5 Suppla.ental Structural Inspection Progra. (SSIP)
As the high-time airplanes in the world fleet of jet tr~nsport catego~ airplanes began to approach their or1ginal lifetime design objectives, the industry questioned the continu~ airworthiness of the ~girg fleet !1nce llany of the airplanes would continue in service beyond design objectives. This concern ultillately led to a requiretaent for a structural reassess.ent or audit and the develo~nt of a continuing structural integrity program for older transport airplanes. The air transport airfr .. aanufacturers developed the required programs, utiliZing different concepts, to achieve continued airworthiness of their aging airplanes. The structural integrity programs have re~~lted in directed inspections of SSls (any detail, .1aent, or a$s_ly that contributes significantly to carrying flight, ground, pressure, or control loads and whose fan ure could affect the structural integrity necessary for the safety of the airplane) at appropriate initiating thresholds and repeated intervals to detect fatigue da.age before the lOiS of residual strength of the airplane's structure.

41

In 1978. the portion of 14 CFR 25.571 dealing with fl11-slfe requirellefttswas revised to reflec::t state-of-the-art advances ift fracture ..chantcs and ~tructural analysis. The fteW regulation required conSideration of daaage growth characteristiCS at ~ltiple sites. Ind an inspection prograa to incorporlte these an~lyses to ensure that the daage was detected before tile "Jidual strength of the airplane dropped below the regulatory fln-saf.
requf"..,.ts. <_fcb
lie"

This"$ called

the cSiIIage tolerance

COftCtpt.

Boeing's approach to the aging fleet p1"Obl_ for the 727/737/147 certificated under Ute pre·1978 14 CFR 25.571 criterta) was to reassess these airplanes using the revised 14 efR 25.571 daage tolerance requ1relents. This reassessment required dete .... ination of residual strength

...

37

presence of mu tip 1e act i ve cracks, extensi ve anal}1si s of crack 1 9r~th rates. and incorporation of these engineering determinat10ns into the airplane's maint~nance program. Boeing applied the same methodology to the 'reassessment of the early model airplanes that was developed to certificate the IIIOdels 757/767 in accordance with revised airworthiness regulations. The development of the program was a cooperative effort between Boeing and an industry steering group. The FAA and the Civil Aviation Authority of the United Kingdom were observers, and the FAA subsequently aaandated implementation of the program by an AD. For the 8-137, the program was to be in effect no later than November 1985. Using a prdbabil ist ic approach which assumed that fatigue crack~ng ~ad occurred ~n the fleet and that the highest time airplanes were the ones that would encounter cracks first, Boeing recommended a cand~date fleet of high-t1me airplanes to be inspected under the SSIP. For the B-137, the candidate fleet consisted of about 125 airplanes, including the accident airplane operated by Aloha Airlines. Positive crack indications were to be reported promptly to 8oe;ng~ where the discrepancy would be evaluated. If the problem was applicabie to the rest of the fleet, an S8 for inspection or repair would be issued and subsequently mandated by the FAA through AD action. Since the program was devised to detect instances of previously unknown fatigue cracking of a structure, the SS! was to be dropped froll! the program once fatigue cracking became known and correct.:;:d through the AD process.

wah the

was devised t~ determine which SSIs ultimat~ly

a structural classification system would be included in the SSIP. Only the SSls where damage detection was to be achieved through planned inspection were included in the SSIP. One of the classifications by which SSIs were excluded from directed suppae.,;:"ta' inspections was that of "duage obvious or ~lfunction evident." An example of a structure that meets this classification is wing skin, whe-re surface cracks are evident through fuel leakage, and fuselage m;nimum gage skin that an~unciates a failure by controlled decOIIIPression throug:-a flapping. Other IAnufacturers include fuselage skin in their structural in:.:oection requirements. Airl ines incorporated the SSIP into the Mintenance airplanes they operated. Among these ai",laHS accident airplane~ and N73712, as stated before, the highest werld fleet.
candidate had

During the program formulation,

Aloha progra:ss of the were N73711, the cycle 137 in the

Aloha "-11"1 ines' incorporat ;on of the SSID progr_ tnto its _intenance schedule was approved by the FAA. The SSID provides the operator ~th procedures to evaluate and suppleMent their existing structural 1ftspection program by utilizing directed supple.ental inspections. A1. Airlines had not discovered or reported any items foll~ng the perfor.ance of SSID inspections. 1..17..1

FAASUrveillance The FAA's

of Aloha AtrUnes llaintenancc Miintenance

Inspector (PMI) has the cOItPliance with Federal regulations

respcmsibility

to oversee an airline's

PrinCipal

with respect to .ainter-ance, preventive ~ainten~nce. and alteration programs. The PMI detenaines ttJe need for and then establ ishes work programs for surveillance and inspection of the a'irl ine to assure adherence to tha applic.able regulations. A portion of the PHI's position descr+pt+on reads as follows: Provides guidance to the a.ssigned air carrier in the development of required .ail.tenance _nual sand reeord '-.ping systems. Reviews and determines adequacy of •• nuals associated with the air carrit~r' s maintenance progY'IJIlS and revisions thereto. Assures that manuals and revisions tom~ly with regulatory requiretllP.nts, prescribe safe practices, and furnish clear and specific tnstruct tons governing maintenance programs. Approves operations specif;c~tions and amendaents thereto. Determines if overhau1 and inspection time limitations revision. warrant

Detenaines if the air carrier's training program lDeets the require.ents of the FARs, is compatible with the maintenance results in trainee and coapetent perso~nel.
progr~, is properly orga~ized and effectively Directs the inspection and surveillance of the air carrier's continuous airworthiness aaintenal'tce progrm. Monitors all ~$es of the air caM"ier's _tntenance operation, including the fo110'1IIing: maintenance, eng'ioeering. quality control, production control. training, and rel iabn ity prog:"UlS.

~onducted, and

At the Safety Board's public heari ng on the ace 1dent, the PMI for Aloha Airlines at the ti.e of the accident stated that he was trained as an FAA air carrier inspector and had been assigned to Aloha Airl1nes since Jauary 1987. He attended a recent course in aaintenance p~ann1ng; howeyer., he ~ ~t received any specific training in corrosion control, .ultiple site fatigue daage, or .uagaent of high ti_ "lead the fleet- aging aircraf1;. He stated that he was not aware of an fAA c.ourse d~yoted specifically to PMI duties.
The Honolulu FAA Fl1ght Standards District Office (FSOO-13) held the FAA clrt1f1cates of Aloha Airlines. The office Work Planning M&n~t Sm(WMS) records ere reviewed fer air-:raft records ~x_inatton$ and spot ucI ...., iaspections accOllP'lished on 1113711 and N73712. for 6 .,nths before tile accident. The review of these TKCll"dS disclosed that .11 requirod MPNS act1.,t1es ~ been acco.p11sbed and that the PMI .alnta1ned • COIIti...-s svrve111uee of Ute a1rline. In addition. the PMI had been

~1

26. 1988. Me record was found nor rt-quired indicating that the PMI --'Mel tIM! $..... repair Oft 113711 "Mch was $~gned off by the Aloha Airlines
111SfKtor
Oft

iafOlWMl

....

skin

crICks

Oft

S·4l

and

S·,IR were

found

Of'

Nn712

on

IiIoveIIber

14. 1987.

39 fAA surveillance of Aloha Airlines maintenance activities was organized around the daily work schedule of the R1J. In a few cases, the PMI visited the Aloha Airl ines maintenance facil ity eitrly in the morning to assess lRintenance practices. In most cues, his visUs took place after the Rjority of the maintenance work had been accomplished. Thus, the PMI prilll3.rily observed completed maintenance actions rather than work tn progress or the. actual condition of airplanes before the start of a repair.
The PMI stated that his hea.vy work.load assignment made frequent visits to obserJ:e Aloha AirHnes maintenance program impossible. The PMI was responsible for nine air carriers and seven repair stations. These carriers ~nd repair stations were spread throughout the Pacific basin and were situated in the People's Republic of China, Taiwan, Hong Kong, Singapore, the Philippifles, and Hawaii. He stated that the travel distan.ces reduced the time available for surveil lance of Ei,ch operator. The PM! stated that he was ·zeroing in on getting the organization [Aloha's maintenance department] up to date, andernized? getting the ~!~"9ram changed to a program that would recognize ~e changes that [had] tak~n place over the years.the PMI stated that he had recognized a ·lack of depth in Aloha management- and was conc2ntrating his efforts at Aloha "jrlines to resolve this issue. Tbe Pf'!iI believed that lmprovingllanagement would also result in illlProvements in Aloha Airlines operational maintenance progrllll.

It

The Aloha Airlines Op@r,ations Specifications for maintenance inspection time intervals in effect at the tiE of the accident was dated August 30. 1982. well before the arrival of the current PMI. A D·check interval of 11,000 hours .as extendei to 15.000 hours by the previous PMl at the request of the operatcr. The time increase was based on the -excellent reliabil ity of the airframe struc.tu~~"e~ndinspections (which) disclosed no significant findings •.•. • The Alohi,l Airline! Maintena.nce Manual containing the D check program'· had been established in 1972. with Aloha Airlines ~rior to his as PHI. He testified at the Safety Board public hearing that otheT -..beTS of the FSDO, including the previous PMI, had informed him th~t Aloha Airl ines was a good operator ilnd tbat there were no problems with the aa.intenance depart.ent. The new PHI stated that he WlS not llade aware of the high-time status of SOlIe Aloha AirHnes aircraft. nor did he receive any info1"llltion regarding the in·service .odel testing of the 8-137 conducted by the aanufacturer in the fall of 1987. ISsignlleftt

Further, the PMI was not faRiliar

"1 •••• not for •• t ty t •• alr'In •• anuat or etl.n •••• but t~. ,. ponatbfllty •• to r._Se. tba .anva' and pro.ptty ad.I •• t' • any pottlOft f. found unacceptable. (I.feroneo •• '""ol't'1,.o.' '."boot. ..part •• nt of tran.portafton. '.dar.t A.".tton Aft'ft •• tt4tthln~ Ord.r "00.9 July IS. 19a5, trt.ptar 6. S.ctfon 4. ".Intaftanet

ttt .... , .a. ••• rator I....... ct.,..

"fI ....

."11"0 ••

.h."

••,.val I.qut, ••• nt•• ,

.... __ ,.""_·.,·.:....· .... '""d

40

1.17.7

Boeing CoaIereial Airplanes CustOMer Visits

As an adjunct to the Aging Fleet Program required by the SSID, •

aafntenance prog~ams.The program also provided Boeing with information on pNbl_ encountered by the otlerators during 1lla1ntenance~ The objectives of thbprognm were to observe the effectiveness of aaintenance progrus, observe the effectiveness of corrosion prevention ;nd control, gather infol"llation· to ensure safe and economic operation of aging airplan~s, and p1"mKJte improved design of new airplanes. Thirty-five operators frem 19 countries initially were selected for and ag.eed to host tea visits.
The

Boeing initiated a program to assess aging airplane structuresandsyst2lls. The Boeing Aging Fleet Evaluation Program consisted of Boeing survey teams vistUngoperators to assess the condition of aging 701.720, 721, 737, and 141 airplanes by observing selected airplane structures, systems, and

t.i_ airplanes if. the 8-131 fieetand the fact that several of Aloha Airlines 8-131s bad exceeded 15 percent of the airplane's design life cbjectives. The Boeing team's first visH to Aloha AirHnes uintenance facility occurred from September 17 to 23, 1987. During this visit the tea surveyed N131lZ while n was in for a heavy maint~nance inspection. From October 22 toZ9~ 1987~ the team returned to survey N73713.
9

selectifln of Aloha was based on its operation of the higbest flight/cycle

Alot.a Airl ines was one (if the operators

visned

by the Boeing team.

On October 28, 1987s senior Boeing executives met with Aloha Airlines" president. and cbief executive officer and its vice pl"'esident of operations to discuss the fi.1dings of the survey team. At this _eting~ 'Joeing personftel voiced their concern about the corrosion and skin patches fovnd on the two airplanes. At that ti., Boeing personnel rec.-.ended, .ang other things. tbat Aloha Airlines ·put present airplanes down for a period of 30 to 60 days and totally strip and upgrade the structure.· In a letter dated October 27, 1987~ Aloha Airlines requested the ltaintenanceand Ground Cpe:-at ions Syste;;s (MGOS)organization to evaluate Aloha Airlines .aintenance operations. According to Aloha Airlines I8ft1g811ftt staff~ tbe requRst was generated by their concern to upgrlde and lIOdemize their aaiRtenanee progrlDl. A Boeing teaAI visited Aloha Airl ines facilities and ev51uatedits s.;.intenance progr_ in Novetaber 1987.
Boeing

• ..

JMuary 8 to 15. 1988.

.5-4

A siaUar

bod.Y

tiM, the Scaing teaa observed the repair of skin lap spllc.while the airplane was in for heavy .. intenance.

-aging

Att~t

fleet-

survey

of ..,3717

was accOIIIIPHshed fro.

A1,.U,,"

.,.nd,.

1M tIGOS report

I) ..

~I"J'

30,

1988, ad

on uintena,..;e operations was delivered to Aloha contained 37 reca.endations. (See

Oft Apy1114, 1988, Aloha Airlines lilt witb 80eincJ to discuss the ffDdf. of u.. agill9 .ircraft soney tea &nd the MGOSrec~ndaUons with Alolta Atrl1Diu~.. Boeing personnel stated they _re undertbe
_N$$iOft

UAt

AloN

Airlines was pluming

to

delay the

re<:C8ended

41 high-ti.e airplane (N13712) was in the hangar for heavy maintenance. Boeing personnel requested that the FAA PMI be excluded from this meeting in order to ·protect the confidential relationship existing between Boeing and the custOller airlines.· The following recommendations were ude by Boeing to Reinstate plan to conduct complete structural inspection on at least the following airplanes: N73111, N73712, N73713, "73717. Conduct a detailed $-4 lap splice inspection on all airplanes having over 40,000 flight cycles and perform total corrective action on any discrepancies found. Initiate belly skin replacement program. Reinstate existing cnrrosion control program immediately. Initiates when available~
including
Review

structural

overhaul

of its

high-time

airplanes.

In fact,

at that

time~ a

Aloha:

recommended corrosion

Boeing developed maintenance progr .. control program.

and correct? as necessary, supplemental structural inspection prograa and airplane sampling progr .. require.ents.

A package of briefi~ Boeing team visit~ 2r~ briefings

notes and related .aterial pertaining to the to Aloha was revi2Wed by the Safety Board.

It

Board

After the accident~ Aloha Airlines responded directly to the Safety with co.ents addressing the Boeing visits and its doeUllentatlon. Regarding the aatntenanee organization evaluation, Aloha Airlines stated that they bad initi~ted act lens to ca.ply with Rny Df the reca.endattons before tile, received the report. Their reply 1n part satd: Prior to the issuance of the Januilry 30, 1988~ lII.intenance organizational evaluation,. IAn)" of the reca.endations had already been illplsented. Since that date and prior to April 28,. 1988~ several lIIjor progrlltS. including the total reorganization of the Quality Assurance and Maintenance depart_nts, have been acca.pUshed. The raaain1ng ~tions, including a new heavy .. intemlnce progr_ c:urteRtly being written by heing~ Ire in the process of being i..,ll11ented. This progr_ will tailor Aloha's current COf'nisi01' control practices to Boeing's reco_nded c01TOsion

contro 1 procedures.

Since the Boeing report was wrttten. Aloha had added a Staff 'ice President of Quality Assurance and Engineering. a Director of Qual tty Assurance :md a Chief Inspector. These positions were addf!d t:l' assun assertiveness and statUTe of Aloha's inspectors. In ad,Ution. the Manager of Operations Stanct.rd$ 1$ preparing a new Traihing Manual, wh1cb does i.dude a spechl aphasis on corrosion detection. Aloh bas

42

asked Boeing to provide additioNal training for inspectors .... non-destruci.ive testing techniques .nd proceduns.
On March 1. I988.AJoha increased tbe number ofaanagement people In the QuaHty Assurance Il@partllantfrom 1 te 4 and created a new three (3) person operations department for tn1n109 and tec.hnicalpublications and a new manager of shops position. An additional manager and supervisor in ~inten.nce haft been added.

develo~nt

UDder the new Maintenance programt ·C· checks will be KCOIIPHshed tn a two-week extended visit rather than in overnight segllents. As part of the new ·Clt check package, crttical flow charts win be developed to lIIOnitor the

of the checks.

Regarding the aging aircraft evaluationt Aloha Airl ines noted that Seeing did not present their briefing (and slide presentation) untl1 as IlUch as 6.onths after their initial visits. Aloha further stated:
Boeing' s reference to -the deteriorated condition of high cycle 737's- is illustrative. Whendiscussfng the condition of these aircraft followfrog the e__ ncement of Boeing's aging fleet ualysis with (officials) of Boeing. Aloha's President and Vice President of Opera.tionswere assured that the ,ireraft were s~fe to continue in operation. This teleconference occurred on OCtober l6t 1987. These assurances were gfven during Boeing·s aging fleet evaluation. . Like RAY a1rlines, Aloha had relied upon FAA designated eaglneering representatives and engineers froll Boeing's Custa.er Support ~p to assure structunl integrity. In addition, Boeing '";as _tntained an on-sUe custGller support office aanned by I Boeing c.usta.er service repres.tlthe since 1969" Aloha has an effecttyeprogrur of structural repair. Allrequ1red strvc:tural teraiuting actions have been -=cGllP' isbed. Aloha has 42 years of corrosion control experience In I: harsh envircmaent. In fact, Boeing visited Aloha fOTitS aging fleet "'ysis while two Aloha aircraft
I:nd

Aloha aircraft experienced I: high mRIlber of fHght cycles. 1t is also t..... th&t thosQ s_ aircnft fly at lower altitudes Mel pnsDre differentials than other air clrriers.

were UIICSer90tng sdleduled corrosion control

rePl,ir.

WhOe

1.17..

n. 1at....

1 AYfltton Safet7 rupectlon

Protr-

As part of the FAA's "tional Avtatton S&fety Inspection Progr. (MSJ'),.. nott" published by tJIe FAA Oft April 13. 1987. -Interi. CU1dance forCGftdvc!1J!1g Jndepth Insi)eCUans-, sutes, -The objective o1indepth . taspKtioos ti to clttlf'lri .... air carrie.,. cOIIplianee with the FARsi inc.1Udtag CCIIIII*I7·,,-ocIdUftS IftClpo11ci.s th&ta1't FAAIP9rovtd,1ild with _ttten FAA pidaac._tat'1al.· Caidance ift the fo... of iftspection criteria is provided

43

to focus on operational and airworthiness regulatory items which ca:n clearly be recognized as in place or not present. Neither quality assessment of the various programs nor the identification of systemic deficiencies of airline operations or FAA surveillance are included as objectives of the NASIP. N13711 accident was conducted in December Airlines was suggested for inclusion in regi?nal director because the airline inspection. FAA personnel ~ including the

The most recent special FAA inspection of Aloha Airlines before the

were assembled from FAA regions

headquarters.

other

than FSOO-13.

1987 as part of the NASIP. Aloha the NASH' schedule by the FAA/s had nDt had a recent indepth team leader for this inspection,

A preinspection NASIP team briefing was conducted at FAA At the Safety Board's public hearing, the manager of the Flight Standards Evaluation Staff of the FAA stated: There were no items
time--and this briefing

that

were specific to
was

conducted in September 87. Trending up to that point demanded that we look at management~ that we look at Airworthiness Directives compliance, that we look at training programs. those kinds of things.

conducted--for

Aloha at

Aloha.

that

was

FAA personnel

instructed

to:

involved

in the inspection

later

revealed

that

the team was

conduct a thorough records review, look carefully at the airline methods of compliance with regulatory items such as the [lIrinilDUlDequipment lists] and ADs, and then go to the airplanes to insure that things were actually accomplished.
operating ~ircTaft. regulatory

There was no advance inspectiort emphasis placed on the environment, the SSID program, -fleet leader- aircraft, or specific condition of the aircraft on the ramp.

harsh
aging

findings:

The December 1987 NASIP report of Aloha Airlines contained numerous compliance findings. The following wer~ the general introductory Aloha Airlines Maintenance Management has been remiss in their responsibilities by not being able to recognize their own deficiencies as this report will indicate. The size and characteristics of Domestic and Flag Carriers demand a formal Continuing Analysis and Surveillance, Reliability Programs, development and control of its policy ltanual ~ rect'rdkeeping systems and coepl hnce with
andatory areas such as

_nagnent
over

organization

to establish

and maintain controls

its opeTations specificltions~

•
! c

the knowledge and expertise to perform the technical tasks conducive for the airline function. HQwever it will be shown
9

This inspection reveals that the present ta:lnagement group has

~.

i
L

~' .

44

throughout the report, that TSAA [Aloha] management has fallen short of being able to accomplish its obligation for compliance of the FAR's particularly in the area of Continuing Analysis and Surveillance, and Maintenance Reliability. A selection of representative NASIP findings follows:
2.2.4

41

Operations Specifications, Page 11 of 15 authorizes -Reliability Programs" based on the following:

and Whitney JT80-7, dated July Z, 1971.

i.

Propulsion System Reliability Control Program Pratt JT80-9, and JT8D·17, document

b. c.

Auxiliary Power Unit Reliability Program Airesearch GTe P-SS-129 document. dated March 16, 1972. Hydraulic System Internal leakage Test Program document, dated July 20, 1976.
{HIST}

This Operations Specifications page is invalid due to the nonexistence of the aforementioned documents. In ad1ition. the above documents are not on fHe at the Honolulu FAA FSDO.
2.4.1.

Training rer.ords for Aloha Airlines Inspectors do not conta ina desert pt i on or source of the !Aterial used for training in non-destructhe testing. training entries. records fnr 3 SupervisGrs
have

2.4.2.

no

2.4.3

Examination of processes used in maintenance revealed that composite material repair is being accomplished by untrained mechanics. Review of the training program and discussion with individuals verified the fact that Aloha has no training progr~ for composite material repair. Tbe Aloha Training Manual states that a .ini.um average grade of 7~ .,st be attained on all fOnAl clusrOOlt training. In a conference with the person who schedules triining and .ai~ta1ns training records, he stated ~hat written ex_minat ions are not conducted; therefore there are no procedures fo;" grading of training rtceived as required by Aloha Airlines unual.

2.4.4

45

2.14.1

Aloha Airlines, Inc. Continuing Analysis and Surveillance Program ooes not contain adequate procedures and standards to meet the requirements of FAR 121.373 for such a program. The type of finding(s) in this NASIP report serve to substantiate that Aloha Airlines does not have an effective internal audit program. Aloha Airlines, Inc. has no procedure to classify repairs as major or minor and has no information concerning any minor or major repairs» in their General Maintenance Manual. The Aloha Airlines General Maintenance Manual (GMM), Section 3-31. Authority For Change states, "All major repairs and alterations wh i ch are not covered by manufacturer's approved data shall require FAA approval." Thi s ts incorrect because the manufacturer's data must be FAA approved.

2.16.7

2.16.8

After the inspection, the FAA provided Aloha Airlines with a copy of the investigative team's findings and the airline was provided an opportunity to respond. The Aloha Airl ines response was evaluated by the local FSDO staff. Consistent with the FAA Flight Standa~ds policy, th~ NASIP team inspectors were not involved in the f~110wup, review, or closeout of any negative findings. If a response was considered to be adequate by the local FAA staff9 the investigative team's finding was classified as ·closed· and removed from the report. This evaluation process was repeated monthly by the local FAA staff and resulted in a ·Status of Findings/Corrective Action· letter to Aloha Airlines dated April 21. 1988. This letter was reviewed by the Safety Board. It was found that the outstanding corrective actions did not address specific airplane structural maintenance pertaining to thp. accident airplane. 1.17.9
Subsequer:t fAA Actton

188·09-51 applicable to all B-737's with more than 55,000 landings, requiring

The day following the Aloha Airlines

accident, the FAA issued AD

flight at reduced cabin pressure and visual inspections of the lap splices at $-4 and -tOl and R and all circumferential splices between BS 360 and 1016.

After the receipt of more information, AD T88-09-51 was superseded by AD T88·10-51 issued May 4, 1988t applicable to all 8-737's with more than 30,.000 landings, requiring visual inspections of all lap splices and eddy current inspections of lap splices at S-4 and -tOl and R. Additionally, a reporting requirement was included for positive indications of cracking or corrosion. The FAA received reports from 18 operators reporting a total of 49 findings of corrosion or minor cracking (small, isolated cracks). Fourteen a1rplanes had multiple site cracking (cracks emanating from six or .ore adjacent f~steners) .

46

A postaccident evaluation of Aloha Airlines was conducted May 7 through IS, 1988, by a special FAA team from the Western-Pacific Region. The team was instructed to conduct an in-depth inspection of the Aloha Airlines structural inspection program and review the compliance with structural airworthiness directives_ A hands-on inspection of th~ Aloha Airlines fleet

was not undertaken.

AD 88-22-1I became effective on November 21, 1988, requiring the inspections of lap splices and tear straps; AD 88-22-12 was effective on December 1, 1988, reGuiring the inspections of bonded doublers and circumferential spl ices. In addition, the mandatory replacement of the rivets along the upper rows of the lap splices was proposed by a new Notice of Proposed Rulemaking {NPRH} signed on October 27, 1988. The final rule was published as AD 89-09-03 and became effective May 8, 1989.

As a result of an FAA sponsoreJ "Conference On Older Airplanes· held in June 1988, an airline industry task force led by the Airline Transport Association, has recommended to the FAA a modification or replacement program to assure the ~irworthiness of older aircraft. The task force involved some ISO international experts representin9 the airl ines, airframe manufacturers, regulatory agencies, National Aeronautics and Space Adm1nfstr~tfon, and professional aviation mechanics from the United States and Asian and European nations. As a result of the industry effort, early models of Boeing 727, 737 and 747 a i rp1anes will undergo intens ifi ed lIa ntenance and inspect i on i procedures, many of which will require modification or replacement of selected areas or parts rather than continued" inspection. Aircraft arees affected will include lap jOints and bonded joints which have experienced delamination or torrosion. In addition, widesc.1e modiflcation, replacement of aircraft structural materials, fittings and skin has been recommended on the basiS of service experience. Many of the changes (terminating actions) are already being accomplished at airline maintenance bases. Boeing consolidated all of the proposed lKIdifications into a single document for aach airplane type in March 1989. The FAA issued an "PAM for each airplane type which proposes the mandatory c~let1on of the modifications list~d in the 80eing documents when an airplane reaches its econOilic design goal, or within 4 years, whichever occurs liter. Other airplane manufacturers' older .odel airplanes are also under review with similar proposal~ for consolidated service documents.
t

In addition~ the FAA Flight Standards Service (reated an ongoing Aging fleet evaluation teams were forwed with speCialists, district office inspectors, and certiffcttfon engineers. Tnese te&mt are v1$iting airlines to evaluateJ through over-the-shoulder inspection, the effectiveness of the airline's corrosion control programs, $tructural inspection techniques, and AD accompl is_nt. The goal of the Agiag Fleet Program is to recommend methods, policy, or regulatory changes to 111Prove the maintenance program for operators of Ig1ng flt:'etaircraft to ensure that each operator is aware of and is applying Maximum effort in t~e Ipplfcat10n of structural inspection progr.ms to allow aging fleet airplanes to continue safely in revenue service.

Aging Fleet Program.

47
2. 2 .1 ANALYSIS

General

The fHghtcrew of flight 243 was qualified in accordance with appl1cable Federal Aviation RegulaUons and company policy and procedures. The airplane was ce~tiflcated, equipped, and operated ~ccording to applicable regulations. Meteorological conditions we,,'e not a factor in this accident. Aerodrome, navigation, and communicat~ons facilities did not contribute to the accident. The Safety Board determined that the accident sequence initiated with the structural separation of the pressurized fuselage skin. As a result of this separation, an explosive decompressi~n occurred, and a large portion of the airplane cabin structure comprising the upper portion of section 43 was lost. The Safety Board's analysis of this accident included an evaluation of the structural and metallurgical evidence to determine the initial failure origin and the manner of fuselage separation. Further, the Safety Board analyzed the quality and effectiveness of Aloha Airline's maintenance practices and the FAA's oversight of that program. Also, human factors aspects of airline maintenance and inspection programs were examined to determine if important but repetitive tasks can be performed more accurately by the assigned personnel. The Safety Board also evaluated the B-737 structural design and certification concepts and the support role of the manufacturer and the FAA regarding the continuing airworthiness of high time/high cycle 8-737s specificallYt and the "aging aircraft" fleet in general. Finally. due to concerns about the continuing airworthiness of aging transport category aircraft under existing pelicies, practices, and regulatior.s, the report analyzes the existing design concepts and regulations that permit a transport aircraft to have an indefinlte service life based on proper maintenance, inspection and repair. 2.2 Origin of FuselageSeparation A postaccident examination of N13711 revealed that the remaining structure did not contain the or1gin of the failure. Since the sea and air search did not locate recoverable structure from the airplane, it was necessary to determine the failure origin by examining and analyzing the remaining structure and the airworthiness history of the airplane. An examination of the production butt joint at 8S 360 (the forward edge of section 43) revealed that the frame was intact, as were the skin and rivets forward of the Joint centerline. Aft of the jOint centerline, nearly all of the rivets rema;ned in the splice doubler. These rivets were deformed aft, although the rivets between S-SL and S-4R· were also deformed to the right of aft. This deformation indicated that the skin immediately aft of as 360 was intact up to the time of separation. Similarly, the skin at the butt joint at BS 540 (aft edge of section 43) was intact at the time of the

48 separation. as indicated by the forward deformation of the rivets in the splice doubler. The BS 540 frame itself also was intact. This evidence indicated that the primary failure had originated at a location between BS 360 and BS 540 and that the skin at these butt jo;nts was pulled away in tension overload as a result of the primary failure. •

As a result, the areas along the longitudinal separation of section 43 were examined as a likely area for the origin of the failure. Very little of the structure in section 43 from the left side of the fuselage above the floor was found. The frames between BS 360 and 540 on the left side had broken at floor level with a substantial portion of the structure sepa~ating outward, downwardt and aft. This mode of separation was corroborated by the degree of ingestion damage to t~e left engine and leading edge damage to the left wing and horizontal stabilizer.
In addition. the right side of the fuselage portion of section 43 that remained with the airplane was severely distorted and bent outward more than 900• f;ve consecutive floor beams at BS 420, 440, 460, 480, and 500 were broken all the way through. Also, adjacent floor beams at BS 400 and 8S SOOA were cracked nearly all the way thr~ugh. Most of the center floor panels on the left side from BS 360 to BS 947 had lifted. The right side cabin floor panels were not displaced, and little if any distress had occurred at the fastener locations for these panels; however, floor panels on the left side of the cabin between 8S 400 and 8S 500 along the inboard seat track were displaced. This damage sugge;ted that the initial failur~ was on the left side of the fuselage. Further, the size and the characterist"lcs of the separated area, coupled with the intact structure at BS 360 and BS 540, indicated that th£ defect was oriented longitudinally along the fuselage. The severely damaged left cabin floor suggests that the origin of the initial failure area was on the left side of the fuselage. As the cabin pressure in the upper lobe was released, the pressure in the lower lobe was contained by the cabin floor. However, the cabin floor was not designed to sustain a large pressure differential. Consequently, the cabin floor of N13111 deflected upward during decompression, and floor panel failures allowed release of the pressure in the lower lobe. Studies conducted by Douglas Aircraft following a foreign operated DC-lO ac:ident'1 related to a cargo door failure disclosed that distribution of the pr@ssure differential during an explosive decompression peaks at the point of the opening in the fuselage. This pressure peak also can cause the maximum damage to the floor, depending on the strength of the floor structure, the magnitude of the pressure differential that exists before the failure, and the size of the opening in the fuselage.

11Turkfsh
March 3.1974.

Airlines

Accident.

Dc-tO.

TC·JAV.

Er.ononvftl.

F~r.st.

France.

49 The point of maximumfloor deflection on N73711 occurred at and to the left of the inboard seat track for the left side seats at as 440 (seat rQW 3); this ;s an area of the fioor that is strengthened to sustain cabin seat loads. Therefore, the Safety Board concludes that the initial failure occurred on the left side of the fuselage in section 43~ probably near as

440.

Because of the damage pattern on N73711and the service history of the lap joints on earlier 8-737s. the most p1"Obable origin sites were the three upper lobe lap jOints on the left side of the airplane--S-4. S-10. and S·14. The lap jOint at S-4L was eliminated as the location of origin of the failure because of the aft and right movement of the separated structure along as 360 betweP.n S-Sl and S-4R. The movement indicated that the origin wa~ lower and to the left of S-Sl at the lap jOint along S-lOl or S-14l. Had the fuselage first separated along S-14l (below the window line Hne}t there would have been only a small fuselage wall reIIllning above the floor to react to the pressure inside the cabin. Tbe resultant force reacted by this saall wall area and internal pressure would have been insuffiCient to bend the wall outward to break the frames. However. the fuselage on the left side was torn extensively into the lower lobe~ and the fuselage frues had separated above and below the floor line. Thus? the Safety Board concludes that the separaUon was probably above the lap jOint at S-14l. that is. at the lap joint at S-lOl.
and above the floor elNDating

At BS 520 in the relDiining fuselage. there were fatigue cracks longitudinally fTOll both sides ~f at least seven adjacent rivet holes in the skin along the lap jOint at S-lOt. Although this was not the failuTf! origin. such cracking is indicative of the type of preexistent cracUng that probably was present along random areas of the lap Joint at

S-lOl.

Further~ a passenger had noted and later reported a skin crack aft of tbe forward entry door near a top row of lap jotnt rivets for S-lOl while boarding the airplane.(Tbe passenger later was escorted to a si.nar atrplane Md verified the cbseNation.) The Safety Board believes that the tcp rivet TOW was crac.1(ed at the S-lOl lap Joint just aft of as 360 before tile accident fl igbt takeoff. Additionally. paswngers seated Oft the left side of the airplane stated that the location of tbe Iriss1ng flight attendant '.ediatel" before tile cfecoIIpressiOft was in the aisle at seat row S. OUPing the decoepression,. evtdeftce indicates that the flight attendant was ejected f... the airplane at • location corre$pcmding to S"lOl near 8S 440.

. CcmseqvenUy, the Safety 80ard det.raines t... the 'usel. of t 1737U tIOSt pyoobabty failed catastroptttcally along S·10L. initially near IS 440, allowing the upper fuselage to rip ffte. The reason for tMs caustrortttc failure. rather than theintetMled fatl·saf. -flapping- of the skin as designed, was ev.luatec: bl the Safety Ioard ..

2.3

Fuselage Separation Sequence The redirection of a lJftgitudinal fuselage crack, and thus the

success of the flapping IleChanisat to safely decQIIPress an airplane~ depends Oft the integrity of the structure ahead of thQ ":rack tip_ If tear straps are disboaded. they hea.e Ineffective because st.; tening is Iest , aftd the crack can propaga~ as if the tear straps did not exist. In that case. controlled (safe) deccapression aay Nt occur.

Multiple site d1l2ge (MSD) escribes .ultiple fatigue cracks along d MSO Uft range froa a few fatigue cracks uong llany rivet holes to thetlOrSt cue of Dall" Yhully undetectable fatigue cracks _nating rnD both sides of rivet utes along a cDilplete row of skin panel fasteners. It Is theorettcally JIOSsible thit this wont case condition aay result in a QUstrG,'dc failure af tbe fuselage b@foreany crack is visually detected. The ~ of NSD also tends to negate the fail-safe capability of tbe fuselage. Art FAA report (on Ute subject of NSf) is included as appendix J~
&

riftt line.

The ft$O fowel ck>.-ing service on other B-737 lap joints in the worldwide fleet was ta raa1a. area.s along the lap joints. The initial eee.. rence of dtsbondecl lap Joints in randca locations leads to fatigue cracttng in raRdoa ir&a.5. ."specUoas and ex_inations of the re.alning portion of the acddut airp'lue 1. the upper lobe aft of the fuselage ,.,.a.Uoa Uta revealed that tILl MSD ~ .. prevalent tn the aid-bay areas AS (betwen Hjattftt cin;aferenti..ll tear straps). FaUgue cracks up to O~53Inch tA leagtJlwre evident in the lap joiftt along S-JOl neal" as 520. The section of S-4R recovered tbtt right wing leMing edge contained DIIDIrOUS faU.,. cracts tJtat su.e1 f". disboading of tile cold-bonded lap

f,,_

joiat
'attt-

and dtsboGding

of the hot~bond£~ tear straps.

cOllCelltnti_ tAt leads to fatigue crlCiittg ",-. tile rivet bole. Therefore, _11. & disbonded lap Joint can lrith$bnQ. the pntssUI"ization cycles that a P'toper17 bMrMd Jo'nt is irrtad«d to carry,. tfle lip jOint beca.s _re susceptible to fa.Upe crKking.

De rims ast .carry tile load. CG1II1tenunt,. tile btle edge cmted

tile i"tended function of tl. bond (to carry the hoop stress and loads ~ the lap jot.t) h lost __ the joint is disbonded and

bJ

~$.

tile cOUAters1nt procktces a stress

the

lap joint

rivets

are

It is "..1. that _rou .. n flt.gue cTacks .long S-lOl .101'" to fon .1.". cnct. (or cracts) s•.t!ar to U. cract ~t S-JOl that tM PlSS4fl9lr s_ ..... boardi.., tN accideftt fn"t. The d ... ge discovered .. tAt ICC't4Mt ge Oft otIaIr ai~l ... s in the Alohi AirH,.s 111ft. f.1.... stn.U.·grwttt rat.s, .... VIe s,ntce btstol"7 tlte B·137 of 1. Joiat 41t'" ,,,.1_ ltd tM Saf", Board to conclude that, It the ti_ ..,0. fon 'atgw crlCb iQ. tM 'us.l.,. skin 1.p Joint along $~1CII.11... up .tal1 to eM' CIUstl"Opllic f.n... of a large secU. of

.,,,lIM, •••

. st,..

.. .,... Stfd.:r .. ", btUDU tUI. mfte'at cractfng or telr 4ft,.,.. (or I Ulll!lbiHUOft of IJoUlJ eatstld ift !1Mt lap joint at S..lOl to .... t. U. desJ_-iDtnded cootrolled ~$sjOft l'tf the structure.

*,..

tee"".•••
1."

,'lti.,.

51

The Safety Soard further believes that Aloha Airlines had sufficient information regarding la,p Joint problems \.0 have implemented a maintenance program to detect and repair the lap joint damage. The information available to Aloha Airl mes on lap jOint problems included the following:

o o
o

the 8-737s in the Aloha Air1ines' fleet were high-cycle airplanes accumulating cycles at a faster
rate than any other operator; operated Aloha Airlines environment; in

harsh

corrosion

Aloha Airlines previously eract( along lap joint airplane;

had discovered a 7.S-inch S·lOl on another 8-737

Boeing had issued~ and records indicate that Aloha Airlines was aware of~ a S8 covering lap Joint inspectio~ and repair in 1972i revised in 1974, and upgraded to an AS8 in 1987; and

the FAA had issued an AD in 1987 requiring inspections of the lap joints along 5-4 and referencing the Boeing AS8, which called for inspection of other lap joint locations, including along S-10. Maintenance Program

2.4

Aloha Airlines

The Safety Board identified three factors of concern 1n the Aloha Atrlines lliaintenance program. They were~ a high 'accumulation of flight cycles between structural inspections, an extended time period between inspections that allowed the related effects of lap joint disbond, corrosion,
and

structural

fatigue

to accumulate, and the manner in which a highly inspection program was implemented.

segmented

The Aloha Airlines stractuYal 0 check inspection interval for the cont1nuhag afnrorthiness of their 8-731 fleet was approved by the FAA at lS,OGO hour'S. lhe selection of IS,OOO hours appears to have been more (FlurYattve than the 20,.000-l\our interval reea.ended by Boeing. However~ "'cause of the daily frequency of short duration flights, the rate of acc_lat1on of fliglltcyc1es :.n Aloha Airlines airplanes exceedec! the rate .teb Boeing forecast ""en the 8·731 "PO was created. Aloha Airl inesrecords of aircraft uti 1ization indicated that theiratrplanes accumulated about tltrte c1Cles for each hour in service. The Boeing economic design life projections were' based on accUllUlat1ng about 1 1/2 cycles per flight hour. lMs, AlcnaAirHnes airplanes were accumulating flight, cycles at twice the rate for Ifbich UMt Boeing: MPDwas designed. Even with an adjustment for partial pressurizationcytles on short flights, and thus partial loading of tbe fuselage, the ac~UltUlation of cycles on Aloha Airlines airplanes remained

52

and Boeing/s assumptions in developing the MPD.

high and continued

to outpace the other B-737 airplanes

in the world fleet

The Aloha Airlines maintenance program did not adequately recognize and cons ider the effect of the rap id accumu1at i on of f1 ight cycles. The Safety Board notes that flight cycles are the dominant concern in the development of fatigue cracking in pressurized fuselages and the accumulation of damage as a result of flight and landing loads. The Aloha Airlines maintenance program allowed one and one half times the number of f1 ight cycles to accumulate on an airplane before the appropriate inspection. The Safety Board bel ieves Aloha Airl ines created a f1 ight-hour based structural maintenance program without sufficient regard to flight cycle accumulation. The Boeing HPO assumed a 6- to a-year interval for a complete 0 check cycle, and the Aloha Airlines 0 check maintenance program required S years to complete a 0 check cycle. The Safety Board believes that the 8 year inspection intervals in the Aloha Airlines maintenance program was too lengthy to permit early detection. of disbond related corrosion, to allow damage repair, and to implement corrosion control/prevention with the maximum use of inhibiting agents. Of additional concern to the Safety Board was Aloha Airlines' practice of inspecting the airplane in sma)l incre~ents. The Aloha Airlines o check inspection of the 8-737 fleet was covered in 52 independent work packages. limited areas of the airplane were inspected during each work package and this practice precluded a comprehensive assessment of the overall structural condition of the airplane. The Safety Board bel ieves that the use of 52 blockS/independent work packages is an inappropriate way to assess the overall condition of an atrp'l ane and effect comprehensive repairs because of the potential for air carriers to hurry checks in order to keep airplanes in service. Further, the fact that the FAAfound this practice to be acceptable without analysis is a matter of serjoos concern. The effectiveness of Aloha Airlines inspectlon programs was further limited by time and manpower constraints and inadequate work planning methods. Maintenance schedul in9 practices util tzed the overnight nonflying periods to accomplish B checks which, 1n reality, included portions of the C and 0 check items •. However, since there were usually no spare a~rplanes in the fleet, it was obvious to both the maintenance and inspection personnel that each airplane would be needed in a fully operational status to meet the next day's flying schedule, Thus, only a few hours were available during each 24 hour period to complete Bt C and D inspection items and to perform any related or unscheduled maintenance on the airplane. The Safety Board bel ieves that the FAA should include in its procedures for the approval of airline maintenance programs, deviations in airplane use by the operator as compared to the manufacturer's original design estimate, tempered by the operating history of the eXisting fleet. A

41

53

~alendar cap for low-fl ight hour operators and a maximumcyi:le limit for short fl ight operators are more appropri ate inspect ion interval s for these ope,.ators. The Safety Board also believes that the FAAshould reevaluate the criter)i and guidance pr~vid&d to principal inspectors for approvi~g individua.l operator' s maintenaace pl ans that divide structural inspections into a la~ge number of lodependent work packages (segments) to be spread aver the normal C check interval. The Safety Board recognizes the concept that the o check, a.. out l tned in the MPO, for each aircraft is accomplished in a reasonable ~ime period such as 3 to 5 weeks. A true heavy maintenance inspection invol VII!S extensive work which may take several days. Comprehensive structural inspections for aging airplanes, likewise, Cin best be accomplisheC: by a D chec\c in which the entire airplane is inspected and refurbished in 'lne hangar visit. As an alternative, some operators have found it efficie.1t to use yearly block C checks with a phased 1/4 0 check inspection. Any ceviation from this Hfull airplaneH inspection at "season!l schedo1ing intervais" should be evaluated carefully before approval. Operator i .,it i ated changes to mai ntenance manuals and operations specifications are ap~ro~ed by the PHI. Many PMI decisions require knowledge of airplane engineerin£ and human performance far beyond the capabil ities of anyone individual. The Safety Board ,.bel ieves that the PMI should be required to seek additic~al assistance or input from other divisions of the FAA and, through channe h. from the manufacturer and other operators. The types cf input, the sources for both airworthiness and flight standards information and the conditiuns under which such input should be used, need to be reviewed and guidance developed by the FAA so the PMI can perform his duties more effectively. Th~refore, the Safety Bo~rd believes that the FAA should develop and provide gl.:idance to the PHI for the approval of airline maintenance plans which are modified significantly from that outlined in the MPO. ~.4.1 Effectiveness of Inspec'~.ions

An examination of the remaining portion of the S-4R fuselage structure of N73711 indicated that '~he S-4R lap jl)int "ad been inspected and repaired as a result of AD 87-Z1-08 in November 1987. At that time, cracks were detected visually and two repair-s were accomplished. Although Aloha Airlines r4aintcnance personnel stated that an eddy current inspection of the rema-;r;ing rivets in the panel was cencrcted to comply with the requirements of the AD, no mention of this inspedion was found in the maintenance records.

Initial examination of the la~ jOint between the two repairs disclosed visually detectable fatigue craexs that emanated from the flstener holes of the top r~ of rivets. Laboratory examination revealed the presence of many more cracks that were well within the eddy current detectable range. Additionally, it was noted that the upper ~ivet row between the repairs and forward and aft of the repairs still contained the original configuration countersunk rivets .

S4

present in the upper fastener holes of the section outside the repaired area indicated these cracks grew less than 0.620 inch during the time between the inspection in November 1987 and the accident. A total of 2~624 cycles had accumulated on the accident airpllne during this time. After the accident. the cracks ranged in length between O.!10 to 0.154 inch. Ther~fore, at the time of the AD inspection in November, the five cracks ranged from a low of about 0.09 inch to a high of about 0.13 inch. Eddy current inspections perfc1""lD@d Aloha inspectors on N73711 by aftE~ the accident could not detect cracks that were less than 0.08 inch in length, but the Inspect ton reliably detected cracks that were larger than 0.08 inch. Since the striation counts indicated cracks existed in the structure that were above this value (0.08 inch) in length, and that were well within the detectable size for eddy current inspection, such cracks should have been detected along the upper row of rivets in S-4R during the November 1987 inspection. This finding suggests that either the eddy current inspection was not performed in Novelber ~. that the quality of the inspection was such that the cracks were not found.
There are several possibilities why the inspectors, when complying with the AD, faileC: to find the detectable crack in the S-4R lap joint on N73711, even though the area reportedly was given an eddy current inspection and two inspectors perforaed independent visual inspections. Fi/'tt, the human element associated with the visual inspection task: is a f.r.",r. A person can be motivated to do a critical task very wel1 ~ but 1IIhen..s .... to ed perfoY'll that sue task repeatedly, factors such as ~V'''';'.'tion of results, bcredOll, task length, isolation during the in"p ,(': ... task, :1 and the envi--onmental conditions all tend to influence perfoi~"; reliability. that can affect the huaal\ element involved in pertains to the effect of circadiAn rttyt_ on htaan behavior. Airline _intenance is .,st often perfonled at night aMI during the early IIOmtng hours; the Uae of dq that ~a$ bean docUMnt" to cause adverse hUilan perfonaance. Ma1nteMnce prograas {ire .,st effective if tuk schedul ing takes into accour.t t!le possible adverse effKu of sleep loss. irregular wort anJ rest schedules. and circadian factors on the perfonaance of 1Iechani'cs and inspectors.

Striation

COU!1tsof five

of the largest

fatigue

cracks

that

were

uintenance

Another

and inspection

factor

For exuaple. cOIIPliance wittl AD-87·21·08 required a close visull inspection of the lap jo~nts .long S-4l and R and eddy current 1nspectiOft of the upper rov of lap joint rivets along the entire panel in '-teh .feets were found. This illPOsed considerable delNnds on the tnsPtttor if tM results of the inspection were to be reliable. The AD required , ·elos. vfsual 1ns~tfon· of about J,.3OO rtvets and a possibl. tdd1 CU\"'1"eftt in:specUon of about 360 'l"1vets per Pinel. Inspection of the rivets J"ICIUtf'M irJsped.ors to cl Oft scaffolding and _" .long tINt uppel" fuselage car."", a bright 1Igilt wttb th_: 10 the cas.. of an eddy current inspectIon, tM inspectors Neded a probe, 111ft.,., Mel • HgM. At U.s, the iftspedor neriM ropes r.tticbed to the rafters of the Nne,r to ,"vent f.n1ng fro. the lirplM1! .".. it _$ necessary to inspect rivet liMS on top of tM fusel.. Even if the t.... atures were c_"ortable and the lighting H$

i.

55 good. the task

DO

si9"s of .inute cracks while standing on a scaffolding or on top of the

fuselage is very tedious. After ex_ining IIIOre and IDOre rivets and finding cracks. it is ~atural to begin to expect that cracks will not be found.
and psychological

of examining the

area. around one rivet

after

another

for

eva _re difficult. Indeed, the pb:"sical. physiological. 1i.1tations of this task are clear1y apparent.

further. when the skin is covered with several layers of p~1nt the t~sk 15 The difficulties in conducting visUil inspections ~s discussed by Dr.. Colin Drury. a professor of INiustrial Engineering at the State UniYeTs~ty of lew Yc~k at Buffalo. dur1r.~ the Sifety Board's public hearing.

He indicated that in the inspection process,. it is not easy for the huan beif'g to perfo,.. a COftSlstent visual search becali.se (1) the aTea the searcher can coneent~ate on at any one ti.e is li.ited by the conspicuity or size of u. defert to be lookEdfor Hd (2) the search process lilY not be systaut Ie enough; therefore,. !he searcher is PMne to .iss areas that were thought to have been covered. Further, there i:. t"e vigilance decrellent duTing long

-signals· as time goes or. but .:so they give fewer -'a1se detection alarms.Mbat is happening in those cases is that inspectors ~11 change the e~iteria

soebl

inspection

isolaticm.

periods that

Or. Drury te~tified that h~ns

Nve low event rites

and to

tend to

SOlIe

extent involve
f_r

."eet

of' ~t they will report to tM extent tbat an: lncreasingl, 111""=I4!r defect is nteded bef~. tbey win j~ it to b.t reportable. Such vlgiluce dec.--ents occvr du:-iDg very .~0ft9 and Ui)lated inspection duty U_s iA "icb there 1s a. low probanU1 of ftadtnq ~ defect. In such cases. the IM;:Ro being tends to proceed through tM task b7 sa,ylng no ... a decision is to be aade.

.. , JahNKe

.. ~. _rUng nvit'QM:lfttjlo stdfici8t .li.-c.raft cicJNn U_ to perlor. tM tuts U... flutbl. sd*'a1t*l},. Md an undtrstanding of tt. ii!pOf'taac.of Ufeir duties to .. ,... U. ~inort'iMSS .f ttae 11rpl.AH. Al. AtrHMS tnt,.•., J'KOrds ,...,...led tbat lnU. forul training _s prowidld in tl)t bldmtQ.Ut$ Met IMtbods, n. iASpector llIlo found tile S-'R. 1~ j01.lt c:rkU ,.quin", ""if" st_t_ tNt .. 1,. ... t_-Jeb '''liAt., (OJT) hd .... "..., ... slK. 1Ma.c.. .. tM.PKtor ,. AuglRt traiDing nccmIs .. f__ 1 111)1 \r.i .... .,., sept...,. 11 ••• 1 ... r.. ovr b NIS'- was. Ii,," b7 • "iltg ~hth._ hcOris 'nett"t. tINt ,.".tor .... ","14M U. l"fU_' had _,,. 2 Itoun of , ... 1 IIDI tratliwg ..... ,at UNt s.- l ..1W;Ier 11".'.1*9 MS.S tOft 06 $eptIIIbtr 17,. 1911.. prov1dtd b1 1lMfs, .. tM S.fttJ Iomt ~nIJId o.t ... -.c:~ kaN1" tM 1ftSPKtOt' .1'" .., haft possnHIII .tJout db....... COfTOStOftjloand 'aUp cf'Kkf,. at tJW U_ tNt u.,. "'" '_'nd to PIrl"" u. criticil ADifts,.cthlO task. lit f«1. ...... -.os'lton .,. .. tegs" tM .."Sl*t ptrf~ t.. f'rst KJ ."SJ*UOft .. wn1U ,.lel not .rtJ(.~lal •• 1 should lOOk for .... fM9«tt"9 an for Cortht. st9'S~

Anothr factor tlN,t IriJ.y have affected the PlrfOTRnce of Al.·s iUId tDs~iOft personnel ts relat~ to tM quaUty of support proy1did .., Al. ~.ent to aSS1$t u.s. persons in the .,.rfo .... nc. of their Propt-r traifti .. ~ guidac .... ~ ~ .,.. needM as _11 as

taw.

t.,

'$

on

1.'; .'$ a

"rat.,,,, lot,,,,.

!'

_i,.,l,.
ilt •

_i'f\,t_~ • .a'S. .....

."",fMrli

#1-.n

Also...1.... '

n,

H.ftt:~tt4

""'19 s(",l.

,-.1 .. 'ull vtHfl.U., of its ."...,tt.. Thn~ t'Nt aajority of A1."5 .. 17 .,.tng tk ftigfft. It., c:Oft$.i ......

56

scbedule. Such aircraft ut11 ization tends to drive the schedul ing. and indet!d. the COIiPletion of required llaintenance work. Mechanics and inspector; are forced to perform under time pressure. Further, the intense effort to keep the airplanes flying laY have been $0 strong that the _intenaAce pe1"'SOftllelwere reluctant to keep airplanes in the hangar any longer than ibsolutely necessary. Inadequate guiduce and support fro. Aloha llinigelient to its inspectors was 2vident also Wen the Produc.tion and Planning depart.ent sent to the inspector" S Dail box,. the ,.0 ana sa on the inspection requirellents of the 1. joints ~long S-4 Without further review or technical c~t. lhese doaamts ere cOlllPlic~teds critical to ainlOrthiness, and subject to 1nterpretat tOft u eYideftced by t"te di SagTeellent ~t its content expressed by experts at the Safety Board's public hearing. These docUllents needed bigbeT law1 review and W1"'ittengtiidance is to their disposition before being sent to _intenance for action. Therefore, the Safety Board concludes that 'lof"~' s ~t failed to provide adequate guidance and support to its .atfttenance personnel and this failure contributed directly to the C4u~eof this accide1st. While tM f01'egJOing indicates tha.t there were deficiencies in the traiQ1ag. guidallce, &ad support provided by Aloha Airlines laanagaent to its .. tD:~. pt.l"SOftftel" there an iftditations tMt. bad these deficiencies not existed lad the i~pectiOft task been _11 .fiDed and strvetured properly, U. tftSPKttoe 11'TCr rate would still not ~ve been totally eli.iuted. This .Uef il. ift part!' supported by the repart.s received by the fAA after this ICci4tftt as a nsult of AD-T88-10-11 "ieb required iMpaction ud reports on JIOSiU... iadteaUons of (Tacks and (orro$:iOl\ fOUftd 1ft early IIOdel 8-731 a11"P11OH. 1M fAA ncehecl '9 reports of corrosion or crKks ftOt pnvtously fMnd; 14 ., tllese tnst.Mces iDvol'ftd _ltisite cr~"ing. No utter how _11 orpatzld a COI'TOStOll dwtKttoea ad crack detection P'rOgTO 817 be and .. IIHWr ... dldfutH Md vigihnt 101 tIOrk force. tile ".specUon ,.«us is '''''I_tty sasupUbl. to ,.. ef'T1)rrate .. lhe.....ore~ the Safety INrd t.U..,.." tNt u. tertli_ting actt. of AO-188-10-U" wb.c.t. consisted ., dr1111,., MIt. the Uisttll9 rhUs and ...,ladns tha with protruding .. ~ nftb. .. xt t.. adoPt.. ,...,tl"l., b1 • fttIiItI.IItr of air1 IDeS.W&S supel'~or to u. GOt" t. u. ID tMt ,.,.tUd ntOHted tupect10ns and patcMng C'rKb; • pnKftS tail ,-.lGcl '*\Mt1.1 pot_til for huMft perfor.nce .,..,.... s..f.tJ INri fw117 COMWTS "'til nc_t FAA t •.-.iNting action for n. 1. jot"" ~t ,..qut,.." tasul1atiClft of protruding '-ad riwts as the ....... ..., sol"._ to 0. 1.., jotat dlsllMdtDg ud ',Uv- probl...

tllPOrt.t

that the airplanes

be available

again for the next day's

flying

t...

0'

.... 7 n

ung tM Glfftcull1 in tM of alrli. _tfttMuc. t«soec.1 is tt. fact tMt FAA ~ tt'l'.'J"9 'or .t~T.ft .. iDtIftMC. t..:""'c'"s contlias _t .. 'al ttlat It; 1-.1, 1"ttl.., ... to tM tim tMt ltnoud JMrsoonel ,11111 actuall, ,..,., ... t... . 'il"U" 8Yi, ... frto. For .... 1." 14 ttl 1'7 •• ,eft 9O"ntS 1M ,.,.'une..". of _,~. ptrsaMtl. ,..1"$ that studRts 1ft FAA .,CNM .ittt,QlIIC:. SCIIWM1, .. kMwlldQeabl. itt sudt topics as wood
fluprKttOft

ltw Saf.-t1 IMnI btU......,. t.. t uac ...

taw

at"'''''s... t ... wfI!M

ai-,I,... f"1~ "",.,.. .no 'JM FAA 'JS. c~U'tcaUA9 air

.,hatton of paint
trlMport .ircraft

and dope. In a wit" fl1 by wire

57

technology. composite materials construction and computer self monitoring capabilities~ tne word~ ·comput2r and ·composite" do r.ot appear in the list of required curriculum subjects among airframe systems and conponents in 14 CFR 141, Appendix~. The Safety Board believes that current requirements
d

personnel will actually perform following their licensure. The Safety Board is concerned aoolJt how wen the FAA approved training curricula can address the human performance limitations of a relatively simple visual inspection task when the training that maintenance personnel receive fails to address the basic skills they will be expected to perform on the job. The Safety Board believas that the FAA should examine the regulations governing the certification of aviation maintenance technican schools and the licensure of airframe and powerplant mechanics and revise the regulations to address cQnteaporary developments in airplane maintenance.
: :ch

for training

aviation maintenance parscnnel fail

to address the tasks that

Another area of Safety Board concern arises from the fact that no FAA requirements for formal training or 1 icensing of NOl personnel. The Safety Board is aware that the United Kingdom Civil Aviation Authorities and those in other countries have formally recognized the i~rtance of NOI ski~ls and have required in-depth training, skill ~nstration, licensing and recurrent certification of HOI personnel. While NOI technology and techniques in some industries in the United States are quite ldvanced and personnel certification follows the American Society for Noncfestr.:ctive Testing (ASNT) guide1ines~ the aviation industry has not appli~ sucb advanced techniques or practices. For instance, ir. the current tftYironaent any mechanic, including those designate~ as inspectors, could be assigned to perform detailed and critical HOI inspections on airplanes with Httle or no training and with tools that have :'lot been technologically
there are t'_'roved for
SOlIe t i.e.

Because of its criticality and c~plexitYt the Salety Board believes that the MDI.aintenance function should be reviewed by the FAAwith • view towards requiring fOnlal training~ skill demonstration, IPPnntic.sbi~s. and fcrul licensing and recurrent certificaton for NOI i.spectors.
of inspection personnel was another issue raised during of this accident. The concern was exJ;f'essed about what kinds of characteristics are appropriate to consider when selecting persons to perlo,. an obviously tedious, repetitive task such as a protracted NDI '''spec:tton. lnsped.crs Rorull1 COIle up through the sen10rity ranks. If tile), "'v. tbe desire! knowledge and skills. they bid on the position and are salected for the inspector job on that basts. However to lsk a technically UowllClgeable per$On to perlOI'll an ob'liously tedious and exceedingl, boring list. ratMl" tNn t.o ha.Vf! hi. supervise the q~l.lity of the task, raay not be ., approoriate us. of personnel; persons who have deaonstrated a capability to .... up to supervisory duties MY itOt necessarily perfonA well at ,...UU.1t tasts. 1... light of tIM! critical iI.portance of the aalr'ltenance ift$~tion tas~. deIonstrated by this ac~id¢nt, the Safe~y Board believes that tte FAA s~ov1d sponsor. as part of its re~ently initiated humanfactors progr .. on this subject, reseJrch to detenaine suitable .eans for use by air c.rrie~~ in s@leeting inspector candidates.
tH tnvestigation

Selection

'$

r
2.4.2

58

Finally, the Safety Board is concerned that nondestructive inspections in the aviation industry involve inspection methods that are substantially dependent on human beings performing repetitive and detailed tasks. The Safety Board believes that research is needed to improve upon the methods presently used to examine very large areas or perform a very large number of simi1ar repetitive actions to find very sma1l defects. Inspections of aviation structures involving large areas or numerous repetitive actions should be automated to the extent possible, or other techniques developed to eliminate or minimize the potential errors inherent in human performance of such tasks. Aloha Airlines Corrosion Control The policies, procedures, and organization of Aloha Airlines aircraft maintenance and inspection prc;~am significantly affected the control of corrosion on its airplanes. Accor-ding to airplane maintenance records, lap joint and other areas of corrosion were detected, but corrective action was frequently deferred without recording the basis for such deferrals. Routine inspection task cards contained the "check for corrosion" instruction for specific areas; however, a programatic approach to corrosion prevention and control of the whole airplane was not evident. It appears that even when Aloha Airlines personnel observed corrosion in the lap jOints and tear straps, the significance of the damage and its criticality to lap joint inte9rity~ tear strap fu~ction, and overall airDlane airworthiness was not recognized by the Aloha Airlines inspectors and maintenance managers. This was particularly notewprthy when one considers that Aloha Airlines indicated that SB737-53-1039, Revision 2 (1914). was incorporated in their maintenance plan. The overall condition of the Aloha Airl;nes fleet ;ndicated that pilots and line maintenance personnel came to accept the classic signs of on-going corrosior damage as a normal operating condition. The Safety Board was also concerned about the uncommanded shutdown of the left engine during the accident sequence. The left engine fuel control was found in the "cutoff- position; the control apparently was positioned there by the res idual tension in the intact cable or motion of that cable induced by the cabin floor deflection since the cables are routed through cutouts in the floor beams.

Since the point of maximumupward floor deflection (hence maxi.. cable deflection) was at 8S 440 in the cabin, the actual location of the throttle cible finUreS (in the wing leading edge) seemed an unlikely one. Additionally, the broken cable ends tacked the unravel in9 that ts characteristic of cables that fail in tension overload. When the appropriate cable sections were removed from the airplane and inspected more closely, tbere were indications of corrosion. These observations were confinaed by laboratory ex_ination which concluded that the diamC!ters of many of the individual wires that comprise the cables had been reduced significantly by corrosion daaage. This corrosion likely weakened the cables so that they separzted at a lower than designed load when placed in tension by the displacement of the left side floor beams. The cables of the right engine also exhibited extensive surface carrosion where they were routed through the leading edge of the wing. These cables may ~ave remained ;ntact during the

59

separation sequence only because of the much smaller amount of floor belli deflectton that occurred on the right side of the cabin. The damage to the throttle cables appears much the same as the type of corrosion described in Boeing Service letter (Sl) 737-Sl-76-2-A issued on August 25, 1977. This Sl was issued as a result of the discovery by Aloha Airlines that a carbon steel thrust control cable had corroded and frayed. On1y five of the seven strands of the cable were reported intact. The re.aining five strands wete also corroded, and the corrosion was present on the ent;re length of that portion of the cable routed through the wing leading edge. The Boeing recommended action following this discovery was to replace the carbon steel engine control cables with corrosion resistant stainless steel cables on the production line beginning with production line number S03 which was delivered in September 1977. Boeing recommended thlt operato{~ of existing airplanes replace the ori~inal carbon steel cables on prrduction line numbers 1 through 502 as required. At this date, the number of aircraft modified in accordance with the applicable SL has not been established accurately. laboratory examination of the separated cables from N73711 confirmed that they were the original carbon steel ~ype. The Safety Board is concerned that Aloha Airlines did not take advantage of the aanufacturer's corrective action for thes_e cables. especially in light of their initial dis.covery of the problem and recognition of their own harsh operating environment. The record establ ishes that corrosion problems were detected by Aloha aaintenance personnel and, on occasiont repairs were deferred without a full evaluation by management of the airworthiness implications or appropriate reference to the structural repair manual. lhis leads the Safety Board to conclude that economic considerations, a lack of structural understand ing" a i rp13ne ut i 1izat ion. and the 1ack of spare a irp1anes were factors which may have induced Aloha Airlines to allow these deferrals. 2.4.3 En9inee~1ng Ser¥ices At the time of the accident, Aloha Airlines, like un)" sull operators, did not have an engineering depar.tment, Some of the functions that are usually perf01"lDed by engineers at large airlines were accOIIPl \shed by Aloha Airlines Quality Assurance (QA) departaent.
The responsibilities of an airline engineering department generally evaluating and illPlementing manufacturer's S8s and ADs. evaluating airplane accidental or corrosion damage, dl!signing or evaluating repliTs, establishing aircraft maintenance schedule speCifications, and providing technical assistance to other areas of the airline. Another important aspect of engineering staff activities is the oversight of inspector perforaance and related quality assurance activites. include The condition of high cycle 8·737' s in the Aloh .. Atrlines fleet with respect to lap joint corrosion, .ultiple repairs, and and detection of fatigue cracking is an exa~le of what can occur in the absence of regular

60

and knowledgeable evalu~tions of aircraft condition staff.

bv qualified engineering

Aloha Airlines mali.agementcould have recoqr.tzed the illlportance of Alert 58 737-53AI039 in light of their cwn experience with the pY"evious Cfact .10ng the lap joint at S-IOR and could have Inspected all the lap joints called out in the referenced S8 ~hile they acc~3plished the requirements of AD 87-21-08. The same concept appl ies to the Sl recoc:mending replacement of engine control cables which were recognized b}' Aloha as susceptible to corrosion. In addition, a qualifh·j engineer should hav! interpreted the lap joint AD regarding the use of oversize pT~truding head fasteners in the event that fatigue damage was found. More importantly, a comprehensive structural engineering and maintenance program likely would have precluded the deteriorated condition of the airplanes by evaluating and implementing the appropriate corrosion control techniq:Jes and S8s, thus retClin~ng company assets. An additional area of concern to the Safety Board is the extent and skin repairs eVldeJlt on the airplane and the effect that these repairs may have on the damage tolerance properties of the original design_ The acci~nt airplane had over two dozen fuselage repairs; the maj~rity were skin repairs using d~Jbler patches. This concition illustrates the extent to which aging airplanes may continue to be ~epaired (patched) in accordance with existing manufacturer's and FAA requirements.
nUitber of

A large repair or the cumu~ative effects of numerous saall repairs can adversely i~act t~e ability of the structure to contain da~e to the extent necessary to aeet fai1-s~fe or damage toler.nt regulations. Additionally. the structure underlying the repairs can be difficult if not i~$sible to inspect~ whicb can be detriMental where fuselage lap jOints are concerned. These types of evaluations are typically ~yond the expertise of QA and aaintenance departments and must be addreSSed hy qual ffied engineering pers~l. bel1eves that the continued airworthiness of enginHrs ill tbe operator's organization. While the Sifety Board recognizes that Situation .., be econo-ical1y unrtalistic for ~11 operators~ it ~ltevl$ tbat an equiYalent level of safety can be achieved on1y by using enginHrlng representatives fro. SOle other sourc.. Qualified engineers could evaluate s.rvic. infor..ation and afrworthiness directives with particular respect to tM neet aircraft and operating conditions. in. assistanee of 'UMtse 'lAtined eng',,"rs lAY be ayal1able through 1ft industry group or the
airplanes
The

as they

Safety

age would be enhlnt;ed by including qualified

Board

airline operators tbat do not haw a funct 10ning .nglneering ~partaert .. intain a fo,..l altenative to provide .ngineertng services. ..int~nce In $~ry.
dtpart.ent
the Safety Board beHeves did not have suffiCient

_factvrer.

The Saf.ty 8oard. therefore,

reee nds that the FAArequl,. ••

to

thit the Aloha AirH_s ~power. the techn'cal

61

~owledge, or the required programs to me~t Us responsiMHty continued structural integrity of its airplanes.
2.5 FAA Re~lsibil

teensura

the

ittes of Airworthiness Directives

2.5.1

Issuance and Clarity

In-service fatigue cracking in a disbonded area of a 8-737 lap joint was first reported in 1984 (by Aloha Airl ines). Then in April 1987, a foreign operator reported several cases within his f1~t. Boeing acted by revising the existing lap joint disband information, S8 n7~53-i039, Revision 2 (which had advised that prolonged operation with disbanded areas would result in fatigue cracks). upgrading the 59 to Alert status. and notifying the FAA. In October 1987. multiple site cracking was discovered during the .anufacturer's continUed fati9~~ testing of a 8-737 aft body section~ Within the sue time frame, the FAA issued AD 87-Z1-OS which required mandatory fnspection for fatigue cracking. The Safety Board considers it unfortunate that the Boeing Alert sa to inspect all lap joints was not issued after the first instance of cracking, and that the intent of the Alert SS was altered signific •.ntly by the fAA to reduce the scope of the inspection when the AD vas released. The Safety 8o~rd believes that had a full inspection of all lap joints been mandated? the likelihood of this accident occurring lAY have been reduced. Tberefore~ the 1 iatHed AD requirements imposed by the fAA. precluded the continuing .irvorthiness of the aging B-737s and the reduced inspection criteria h; considered a contributing factor to the cause of this accident.
~\t'~f1 Aloha Airlines accOliplished the inspections and repairS associat~c .. "tl1 the AD. they oaitted inspections of lap joints other than UIose alea; 5-4 and they did not replace the remaining fasteners it, the upper row of the S·4R lap Joint with protruding head rivets. as outlined in Boeing ASS 137-53AI039. The AD pertaining to the lap Joint inspections states. in part:

Repair all cracks and tearstrap delaainattons found as a result of the above inspections prior to further f1 tght in .ccorduce with Boeing Alert Service Bulletin 737-53AI09" Revision 3. oat~ AugustZO, or later
FAA·apf)J"Ovee revisions. The

appropriate section of the ASS states, in part: cracks ustng a repair sf.illr to that Rep.ir Manual Subject 53..30·'. all rtIIaining upper tow flush

Repair faUpe
protruding Data~

joint-fasteners in tbat panel Joint ~th

head $oHd fISUM.,..S per Plrt

sHIm in 131 Structural Figure 16, .anct replKe

IY .. Repair

oversized

62 While operators have interpreted the repa'lr instructions listed in the ADnote as requiring the installation of the protruding head rivets as a part of the repair. the FAA personnel stated that its intent was to have protruding head fasteners installed throughout the skin pMel jOint where cracking was found. Repairs of the S-4 joint by Aloha Airlines were accomplished using in the Structural ~epair Manual and excluded replacing the reaaining flush joint-fastners. The Safety Board believes that the 1ft:structions contained in the AD were ine~(act and subject to Irtsinterpretation.
the procedure

Such confusion illustrates the difficulty inherent in attempting to present technical information so that it can be inter~reted properly by the users of the infonution. In the case of th~~ AD, it is believed that the rep.air instructions could have been presented more explicity. This was, in fact. done in subsequent ADs pertaining to the same subject.
disbonding
&lppe~

cieficiency.

The Safety Board is satisfied that the terminating action for the of 8-137 lap jOints and tear straps requiring replacement of the rivet row is an effective measure to correct lchis recognized 8-737

sample f... line n..oor 8-731 airplanes. Fatigue cracks were found emanating fna the fastener holes of a s1981ficant nullber ,)f rivets in the .iddle row of the lap jOint. The Safety Board is concerned that because of the extended lifetime of the 8-137 .ffordedby the tenaiNting action undated for the upper rivet row~ the lower ri..-et on the inner (lower) skin pa.nel eventuall~f will be a location for fatigue eracks to develop. These cracks, if they occur, cannot be detected externally by visual _ans since they are covered by the outer skin panel. The FAA and Boeing should continue to evaluate the early a:Ktel 8-737 airplanes to deteraine the types of inspections, inspection intervals, and corrective actions to be instituted if • significant fatiglle cracking probltlt develops iil the .iddle and lower row ef lap Joint fastener~~.
tbe accideftt airplane revealed anot"ler area of concern wah early

However. laboratory

examination of the S-4R lap joint

"*

2.5.2

_did Research

011

Corrosion control

and 11)1

Mbne it is the responsibl1 it,Y of the operator to develop and illPl.-nt & proper and cOlllPlete Rtnt~nce progr_ appl icable to the .".rating env1ron.nt, the Safety Board believes that the fAA should define kcept.Jble corrosion control progr_ par_ters and provide! thea as a gutde for botb operator and the PM!. The Safety Board believes that an .,.raw!"'s. CGIIIP"'heftSive corrosion control prograa. funy supported by the lllllUfactUNr ad _forced by the FAA" is a critical and necessary step in the antiftUeCI ah_rthiMSS of an aging airplane fleet.

t"

cOn"fli_.

the ,..1. i" ~ragi",safety ....... Uon _thGcls.

for ....

Boarcl also believes tltat the FAAshoul~ ISS.. the lead further research into 1l1Provef! cOn'Os1·)ft detection and fIIny areas of an airplane are difficult to inspect for ley withfn lap joints~ the corrosion Cln go undetected

63

until it is so severe that the damage is evident visibly from the condition of the outer skin. There have bee:'tvarious experiments on NOT inspection methods but none have proved effective ina 11 cases. The NOT equi plllent .anufacturers appear able to suppl)' very technical equipment to other 1nc1liStr1es; however, tn the aviaton industry, neither the IIOst techn1cally advanced and automated equipment nor the human factors involved in using such equipment effectively have been pursued thoroughly. The FAA and aircraft .. intenance interests should challenge the NOT equipment manufacturers with the specific needs for the aviation industry in order to de'I'elop improved, economical, state-of-the-art equipment and to employ methods which minimize human performance inadequacies. Even though the corrosion problems with the carbon steel engine control cables have been known for quite some time, the Safety Board believes that it would be beneficial to once again address this area in light of the c~ble condition en the accident airplane and the fact that some portions of the cables can be difficult to .nspect. The Safety Board believes that the FAA should issue an Airworthiness Directive to the r.perators of the affected 8·137 airplanes advising them of the corrosion potential of carbon steel engine control cables and directing them to the information contained in 137-Sl-76-Z-A regarding cable replacemeJ'lt. 2.. 5.3 FAA OVersight The Safety Board has issued numerous safety recoanendations pert.ining to the surveillance of air carrier maintenance by the FAA. Gefterally,. these safety recoaaendations have addressed lNintenance problellS with specific aircraft. revisions to manuals. and accident or special study identified areas fot' surveillance program improvement. After the Aloha Airlines accident, the Safety Soard probed deeper into the FAA's surveillance IJTOCJr-. including HASIP. The investigation and analysis of the accident revealed sever~l areas of concern including staffing levels and FAA pb11osophy regarding .aintenance surveillance. Tht investigation has revealed that staffing levels in some FSDOs a~ 1nsufficient. the PMI responsible for Aloh~ Airlines indicated th~t he was also assigned a!. tM PMI for nin,'!other operators and seven repair stations thrHgbout the Pacific ri. area;. He also w~s usigre<l out of his
geographtc area

responsibility to participate in a NASIP inspection. The Saflty Board believes that the PMl's workload WIS too extensive for hi. to be ~tely .ffective.
As • r.sult of the FAA sl)Oft$ored Sifety Activity [Yillntton (~jKt WE) tn 1984,. the FAAFlight Standlrds S,st_

0'

Functionll

fuels to inc"l,. the IMIIIber of Itr carrier inspectors. ackIlUOftI.l ,.rsoanel .111 illProft tM staffing sitution, the ~fet1 Board is conc:emed

a s-,..r progr_ st~ization.

to i1llProveinspection guidance. field survenlance, and The FAA bas been Illocated additional hiring authority lad

is now In

....n.

u.
:

dout

It

~~l

k.. utensive

..... qualifications of tile fttWl, hired inspectors and the training of iftspt!'Ctor forc.t. htluse there are a 11.itld nUiber of candidates -.0 air carrier bKtgr0un4s" tht fAA. has hd to hire people with IVlation OT •• litary b.ckgrounds Gt' transfer inspectors fro. general

64

a~tat1on assigments. As a result, the new inspectors are not fully familiar air urrier ma.intenance programs and practices. Although the fAA ,,,y1des a 6·week indoctrination for the new inspectors, it requires several years of on-the-job experience to make the inspectors most effective. Then they progress on a ~reer path that leads toward being appointed as a PHI. There is no specific formal training course for PHIs. Additiofial1y, recurrent training is sporadic and difficult to attain~ resulting in a work force tbit .,st try very bard to sby ahead of the operators and quickly advancing .irenft technology.
with

Increased fAA inspector staffing levels should help with the .anpower require.ents necessary to review paperwork that ensures th!t airlines have cOilpHed with Federal regulations. However, without proper FAA iftSpectiOR of actul airplane conditiont less responsible or knowledgeable operators caB operate airplanes of dubious struc!ural and mechanical i"..-grity. The paperwork "'eview system has beCOllle so entrenched in FAA inspections that an alteration of philosophy is required to create an effective .aintenance surveillance progr~.
ihe Saf~t1 Board sought to identify existitlg boundaries of responsibility of the PMI regarding regulatory coap1 iance and the level of a1nteftance quality daoftstra!ed by the assigned air carrier. Evidence of ICCOURtabH tty of the PlU and district office for the perfonaance of the assigned QlTier(s) was not apparent. Evidence suggests that FAA sUJ"Yeillance and inspection progr~ aTe directed toward tlte air carrier, and u.e fa-bouse evaluation of PMI perforaance is oriented toward quaetity of wort ad tile abilit, to hlndle approvals SIIOOthly and directly. The Safety Board is coacemed tbat the PMI bas the authori ty to approve crit ical areas of afr carrier _iateDance program without being held responsible for those

"""_als. ~l~t".

.... ]I ••llt

TMre does not qpear to be an effecUve _thod in place far FAA to .ue ncurr1ng qAHutive assesSEnts of PMI approvals.

FollCNUp of the NASIP ftndings is also indicative of a lack of PMI The aegaUve fbdings of an airline uintenance progr_ are ,laced ~1Ito tile oversight of the PMI to promte and .anitor con-ec:t1ve actlOft. lbat tsy a oegat... situation.., occur under the jurisdiction and SW\f!.o 111 .... of a "" and yet he is responsible to evaluate and followp on COt.ectl .. actiOft. lMrefore" the accountability for the on-going quality of U. PIlls wri perfoNIIICe does not appear to extst.

I:' appears tIlat tbe current surveillance s1st_ can lead to -rubber and endors_t of aa a1." carrier' $ operations and _11IIt_1ac. prog'lU.. J.. ~AtS are ftHded to encourage and sUPP<.t,'t the ~I's efforts to secu-e ca.pltlnce aad to pro.ote upgraded levels of ..,... _Cl: ., U. .IIPId .. atr carrler in both "fety and reHablliy areas.

st.'C,· ....,.wals

......

.. tllhnsts of tile 11r carrier. It appears the present 5YSt_ " tile pef'SOMl t.tegrity and dedtcaU. of the coacel"nlM fAA t1t:SfJ'llC\.1r 1 ratMr ..... ." 1ft fAA s1sttblt includes ~.te ....-sttltt and tttt 1 rw'IW. IHb tbe curreRt eftYi~t; only tM MOst
bd PIUs .. n ..lAtat. their S8Qst' of ,..sponsibility to ensure IlUt_ _ Hfdy.. n. Saf.t, BMrd recognfz.s the need 'or increased

'"~

U. -'it .. W

s.cb tIIpr ...... fttS. tM $1$t. of

,rogr_

approval

Cion

be driven by

.,ffCfaq

6S

FAA management emphasis on the accountability of a PHI's performance. Both regional and headquarters Flight Standards staff should become more involve1 in assessing and ensuring PHI accountability.
In addition, there ts also a need for a program of standardized approvals of air carrier maintenance programs to proaote a uniform and acceptable level of safety performance in the current competitive air carrier industry. The Safety Board believes that the authority of the PMl for approval of airlines procedures and operations spacificaticns can be better guided, and overall PMI performance improved, if definitive Flight Standards criteria are provided to those in the field. The FAA management's primary tool for overseeing surveillance by the inspector work force, WPMS, requires inspectors to enter the number, type, and res~lts of inspections performed into a computer. During Safety Board interviews, inspectors have expressed concern that this system creates a~inis~rative requirements to the detriment of the time available for performing their surveillance responsibil ities. The Safety Board recognizes that FAA management requires the data for their workload and personnel management systems and that the information is used to ensure that inspectors perform the required inspect ions. However, at present, the information plays only a limited part to enhance the qual ity of airline surveillance. The Safety Board suggests that further improvements can be made to streamline the syste~ and perhaps to gain more qualitative information about both the carrier and the PMI surveillance.

The Safety Board also investigated the effectiveness of the NASIP after the Aloha Airlines accident. A NASIP inspect ion had been pe'l'formed at Aloha Airlines in December 1987 and none of the findings and corrective actions addressed airplane structura1 maintenance. In fact. NASIP looked chiefly at manuals and records with a minimal effort expended to the condition of the fleet. A month earlier, Boeing had performed a ma;ntenance evaluation of the carrier at Aloha Airlines' request. Boeing found several areas of concern including the deteriorated structural condition of the A~oha Airlines" high·cyc1e airplanes and Aloha Airlines' innediate need for a structures engineer. The Boeing inspection provided a convenient yardsth':.k by which the effectiveness of the NASIP effort can be measured. The Boei~l effort concentrated tnitial1y on the actual condition of the airplanes. aDO then it reviewed the paperwork to find out why the maintenance program had resulted in the airplane deterioration. The Safety Board concluded that there are inadequacies in the NASIP objectives and methodology which require a change in the current philosophy of FAA surveillance to include added inspection of fleet airplane c)ndition. The Safety Board also believes that routine surveillance and the be adjusted tow.rd a more -safety-orient(~· qualitative prograa to compl_nt the current -Federal regulation cOIIplhnce- approach. That is. under the current philosophy. the FAA examines airline records for c_Hance with requlations. and some negative findings (vicllattons) result fn enforCeMent actions for which there are clear guidelines. However,.any negattve findings are -nonregulatory- matters for which both the local PHIs and tM HASIP teams be~ieye corrective actions should be taken. In the

NASIP concept should

66

prealRble to the 1987 NASIP report of Aloha Airl i:'les, the FAA team stated, ·Aloha Air1 ines' Maintenance t4anagement has been remiss in their responsibilities by not being able to recognize their own deficiencies, as this report will indicate.-Responsibilities" apparently refers to regulations under which Aloha Airlines is charged with maintaining its ai!"plane in an airworthy condtt, ion (FAR IZL363). -DefiCiencies" in this case apparently refers to items which the FAA believes Aloha Airlines should correct to operate safely. There was nc national FAA program to evaluate and verify the quality of the corrective actions, nor to determine the timeliness of sucb actions. For example, the NASIP team found that Aloha Airlines • ••• does not have an effective internal audit program." Although FAR 121.373 ·Continuing analysis and surveillance" addresses an air carrier's responsibllity to maintain ,1 system for continuing ar.alysis and survefl l ance of its inspection and maintenance programs, the FAA NASIP inspectors apparently concluded that the regulation was too subjective to use as a baSis for enforcement action to assure that Aloha Airl ines corrected deficiencies in. their internal audit prograll'. Technically, as stated by the FAA, if an airline complies with the regulations, it is ·s1lfe. However, many regulations are subjective in nature and are subject to interpretation. Consequently, even with several Significant negative findings by a NASIP team, as was the case with Aloha Airlines. the airline was allowed to continue operations ~ithout making i..ediate changes ilnd with('ut having to set deadlines for completion on reco.Iended actions. In fact, the oversight and closeout of cerrect tve ACtions suggeste<i by the NASIP team were left to the Aloha Airlines PMI, under whose jurisdiction and routine surveillance the discrepancies had been allowed to exist.
U

Airlines ~n January 1988. Although several corrective actions were taken by Aloha Airlines in early 1988 and several others were in process at the time of the ,ccideAt. Aloha Airlines continued flight operations uninterrupted during this period~ despite the negative nature of many of these findings. regulation
The Safety

SolIe negative findings of the FAANASIP team were similar in nature to the findings of the Boeing team that evaluated Aloha Airlines maintenance progr_ in NOvellber 1987 and the -aging neet'" survey tea. that visited Aloha

Soard is concerned that an airline that is charged b)f conduct operations in a certain IIInner can be found in DOnCOIIpliince with the intent of the regulations and yet, it can continue ~rc1al fligbt operations without substantive interruption or corrective actions, taken. If an airline ts either unable or unwilling to develop and _inta'n ... .ffect tve _intenance and Inspect ion prog!"lII, the current FAA overs,ght pbtlosophy .. ,n not prevent deftcincies from occurring, and it will not verify that substantive and ti.1, corrective actions have been laUn. Furl..... re, the FAA did not intend to act~l1y -inspectAloha A'rliftes O .. t to verify if the airplanes were, in fact. safe. The findings of tbe Safety Board fonowing the accident of N73711. regarding the quality of A1. AirUnes _tntananee program, suggests that the FAA routine and special inspection Pl"09r_s were not effective in verifying that. the airplanes. were MintaiRed in a safe, airworthy condition. to

67

2.6

Boeing Boeing 737 Certification

2.6.1.

While the initial certification of the 8-737 conformed to existing regulatory requirements, the accelerated fatigue testing schedule did not compensate for the in-service 2nvironmental effects on the bonded lap joints or tear straps. The inadequacy of the testing schedule was due in part to the state-of-th~-art of laboratory bonding verification testing which had not yet discovered the problem of long term bond durability. The bond production problems did not become known until several years after the airplane went into service. It appears that 80eing addressed this problem swiftly ~hereafter by issuing S8s, improving the bonding process, conducting additional fatigue testing, and ultimately eliminating the cold-bond method and redesigning the lap joint. At the time of the initial certification of the B-737, a consideration for MSOwas not a part of the certification requirements, nor is it required now. This is demonstrated by the fact that there is no specifiC FAArequirement for full-scale fatigue testing to multiple projected service lifetimes of an airplane. Boeing atterr.pted to assure fatigue life by testing the representative half fuselage section to two lifetimes. However, the durability of the lap jOint cold bond appears to be the governing factor producing multiple site fatigue cracking in the 8-737 lap jOints. The Safety Board believes that the Boeing fatigue tests of the fuselage to two lifetimes did not generate fatigue cracking, probably because the lap jOint and tear strap bonds on the test article were initially of good quality. Nonetheless, the Safety Board believes in l,ght of the increased knowledge of and concern for the occurrence of MSO, the difficulties that may be encountered in detecting this type of damage and the catastrophic failure that may result from such damage~ full-scale fatigue testing to a minimum of two projected service lifetimes should be required for certification of new designs. The Safety Board believes that full-scale fatigue testing obviously is not a substitute for a comprehensive structural inspection program throughout the airplane's service life. The effectiveness of these inspection programs as the airplane ages would be enhanced by the early identification of areas where MSO does occur and incorporation of the necessary prevent lYe des ign changes so that MSOis not a sign i fi cant factor during the airplane's operating HfetiM. The Safety Board is also concerned about other Boeing transports, including some 8-721 and early 8-741 airplanes. that util tze the cold bond construcUotl. The Safety Board recognizes that the design of 8-727 and 8-741 airplanes is less susceptibh to fatigue cracking problems in the lap jOints as early in tne service life as those that arose in theB·737 fleet. However. a significant number of these airp~anes are beihg used beyond the KonOllic service 1ife predicted by the manufacturer. The approach to fatigue testing for the early 8-127 and 8-747 airplanes was similar to that performed during the B-737 ceY"'tification in that it did not tnclude the possibility of in service lap joint disbonding. Additionally, although the entire afrplane was tested in each case. only one service life objective was

68

achieved. As a result, the onset point of widespread cracking may not be known. The Safety Board believes that once airplanes of each particular type approach their economic service life. that in·depth analyses are necessary to verify the continued airworthiness of these airplanes. Therefore, the Safety Board believes that the FAA should require all u.s. currently certHicated turbojet transport c:1tegory airplanes, and those airplanes certificated in tile future, receive full scale structural fatigue testing to a minimum of two times the projected economic service life. 2.6.2 Boeing Structyres Classification The Supplemental Structural Inspection Programs (SSIPs) mandated by the FAA vary by concept and implementation from manufacturer tc manufacturer and from model to model. As Boeing devised the SSIP for their existing certificated airplanes, a structural classification system determined which SSIs are included in the supplemental inspections. Because Seeing defined the fuselage skin as "damage obvious or malfuction evident" if it cracks, the fuselage skin was excluded from directed supplemental inspection. Other manufacturers use different criteria and include primary fuselage structure and skin in the structural inspection program. Boeing believes that their current FPA approved inspection program is adequate for detecting lead cracks resulting from MSD before the damage becomes critical. However, the Aloha Airlines accident illustrates that it is possible to. have enough undetected (but technically detectable) damage along a Ti¥et line to negate the controlled decompression mechanism. The Safety Board recommends that the classification of fuselage lIiniarJm gage skin as damage obvious be discontinued and the affected SSIPs be revised accordingly. Additionally, all of the remaining SSIs in the damage obvious category should be reviewed in 1ight of the recent approach for possible inclusion in the SSIP.
2.6.3

Boeing Visits to Aloha Airlines

When Boeing visited Aloha A;rlines as part of tts Aging Fleet Evaluation Program, it expressed concern about the deteriorated condition of the surveyed airplanes (N13711 was not included in the survey). Although Boeing did not infonn Aloha Airlines that the airplanes were unsafe, Boeing beHeved that they made it clear that an unsafe condition could result if corrective action was not taken. While responsibility for determining the operational airworthiness of the aircraft rests primarily with the operator, both the manufacturer and the fAA have a responsibility to verify that conditions do not appear that lead to the loss of continuing airworthiness. Both the manufacturer and the FAA participate in the process with initial certification action, operations recOllllendationsto assist the operator to maintain an airworthy fleet.

specifications

approval

and

continuing mafntenance guidance Ind

The Safety Board agrees that it is important to maintain coanunicaUon between the manufacturer and the operator. The manufacturers

69

require information about airplane performance and conditions in service in order to resolve difficulties and provide corrective actions to operators. Evaluations by the FAA of the condition of airplanes to verify regulatory compliance and enforcement shou1d be performed as an independent oversight measure to ensure that the operator/manufacturer exchange ccnt tnues , When Boeing prevented the Aloha Airl ines pm from attending a meeting at which Boeing presented the results of its evaluation, Boeing stated that it was motivated by a corporate concern to preserve the privacy of communication between operators and the manufacturer. Manufacturers should maintain private customer contacts but they must reserve the option to notify the FAAregarding the aspects of air safety of any individual operator should such a need arise. Boeing did not inform the FAAof the condition of the Aloha Airlines airplanes) nor was it required to by regulation. Boeing determined that an unsafe condition ··could developh in the Aloha Airlines airp1anes that were surveyed if corrective actions were not taken and preferred to discuss the findings exclusively with A10na Airlines. Tit.!e j4 CFR 21.3, I'Reporting of Failures, Malfunctions, and Defects, n is intended to ensure that the FAA becomes aware of service difficulties that are not reported through the existing service difficulty system. The lap joint and tear strap corrosion and potential fatigue cracking problems were well known and corrective action was being formulated. The Safety Board recognizes that although Boeing had no regulatory obligation to report the findings of their survey of Aloha Airlines to the FAA, the Safety Board considers it a potential benefit to report such information. Therefore, the Safety Board believes that the FAAshould evaluate the safety benefits that may be gained from manufacturer's survey information, such as aging aircraft reports, and take appropriate measures to ensure such data remain access ib'l e to the appropriate authorities. The FAA's evaluation should leud to more explicit criteria and guidelines to operators and manufacturers about what information should be reported to the FAAunder the provisions of 14 CFR21.3.

2.7

Operational Considerations
-,

The magnitude of the accident was weli beyond any anticipated emergency scenario. The fHghtcrew'.s act ~ons w~re c,,~si~$tent with simulator training situations which minimizet.he exposure to physiological effects. The flightcrew's success in managing the multiple emergency situations and recovering the aircraft to a safe landing speaks well of their training and ainnanship. The cabin crew also performed in a highly commendable manner when faced with a totally unpredicted event. Their bravery in moving about to reassure the passengers and prepare them for lal'lding was exemplary. The Safety Board reviewed three operational areas: assessment of in-flight structural damage; air/ground emergency communications; and emergency ambulance response.

70 It was apparent from crew interviews and the FOR that a rapid descent was 1nitiated shortly after the explosive decompression. The Safety Board notes that speed brakes and 280 to 290 KIAS were used without first assuring the structura" integrity of the airplane (the cockpit door was missing and sky was visible overhead). The lAS used in the descent, although it minimized the time at altitude, increased the maneuvering loads and subjected the passeng~rs to flailing and windburn from the effect of exposure. The open fuselage break was also subjected to high dynamic pressure from the wind force. The Operators Manualt Emergency Descent procedure (and emergency checklist) states that 1f structural integrity is in doubt, "1 imit airspeed as much as possible and avoid high maneuvering loads. If The Safety Board considers that evaluation of the structural integrity and techniques of emergency descent (target airspeed, configuration chanqes , and maneuvering 'leads) can be critical to the success of further f1 ight. The Safety Board therefore suggests that the FAA issue an Ai)" Ca)"rier Operations Bulletin (ACOB) to review the accident scenario and reiterate the need to assess airplane airworthiness as stated in the operators manual before taking any action that may cause further damage or the breakup of a damaged airframe. In the course of this accident, ATCchanged frequency fo~ primary radio contact with the air-plane during the emergency. Fl ight 234 was transmitting transponder emergency code 7700 and after some difficulty, the crew established contact wHh Maui Tower; Maui Tower was initially not appri sed of the full nature of the emergency or the structural damage. After notification of the emergency, ATCdirected a frequency change to Maui approach control. Later, flight 243 contacted Maui Tower for landing. The Safety 8oardrecognizes the requtreuent for such frequency changes during normal operations; however, Maui Tower received both the emergency transponder code and confirmation of a pressurization emergency before their instructions to change frequency. The Safety Board wishes to rei terate that ATC must make every effort to minimize the workload of a crew during an emergency_ Further, an error during the handoff could result in lost communications and a possible loss of positive traffic control of the emergency aircraft.

The Safety Board is also concerned about the circumstances surrounding the ambulance response at the Maui airport. Had the ambulance service been notified earlier by the control tower and been waiting at the airport when the airplane landed, the seriously injured passengers could have been treated and transported to the hospital 13 minutes sooner. It is incumbent on those persons making a judgement for notification of emergency services to be aware of the circumstances and possibilities of each scenario.

71 3. CONCLUSIONS

31.1

Findings
1.

The fHghtcrew was certificated and qualified fDr the flight and the airplane was dispatched in accordance with company procedures and Federal regulations. Weather was not a factor in this accident.

2.
3.

Although Aloha Airlines operated according to the FAA operating certiHcate and operattons specifications, the quality of AlohaAlrlines maintenance and inspection program was deficient. There was no evidence of preexisting failure or malfunction of the airplane's air conditionii1g, pressurization, pneumatfc,or electrical systems that could have contributed to the fuselage failure. The flightcrew/s use of a target speed of 280-290 KUS and speedbrakes in the descent after the structural separation indicated they did not cQJlsider the appropriate emergency descent checkl ist which states, in part,. that if structural ~ntegrity is in doubt, airspeed should be limited as much as possible and high maneuvering loads should be avoided. The left engine became inoperative because the engine control cables separated due to an increase in cable ter-s1on caused by cabin floor deformation, coupled with corrosion found in the area of cable separation. The fuselage failure initiated in the lap jOint 6110ng S-lOl; thefaflure mechanism was a result of multiple s1te fatigue cracking of the skin adjacent to rivet holes along the 'lap Joint upper rivet row and tear strap disbond which negated the, fail-safe characteristics of the fuselage.

4.

5.

6.

7.

B.

The fatigue cracking initiated from the knife edge associated with the countersunk lap joint rivet holes; the knife edge concentrated stresses that were transferred through the rivets because of lap Joint disbonding.
The disbanding of ]apjoints and tear straps originated frOll manufacturing difficulties encountered with surface preparation and/or bond material processing during the construction of the airplane which resulted in lap jOint bonds with low environmental durabtl fty or alack of bondfng. tested to 150~OOO cycles not reflect the fatigue fuselage section of a Boeing 737 was during certification, the test did performance of the actual fleet

9.

10. Although a representative

72 aircraft because the t~st parameters did not consider the long term effects of disbondir.g, corros ton, and fatigue cracking in the lap joints as experienced in airline service.
11.

Disbonding of 8-737 lap joints, w1thresulting corrosion and probable fatigue cracking, was. f!xplicitly defined in Boeing ServiceBul ietin 737-53-1039, Revision 2 dated February 8, 1974; however the serieus implications of multiple site damage were not real ized, a pern:anentso 1ut i on was not determined, and corrective action was relegated to repetitive visual inspections and damage repair. There was sufficient information avai Iable to Aloha Airlines to alert 1t to the cracking problems associated with the deterioration of lap Jotnt bonds, and Aloha should have followed a maintenance program to detect and repair cracking before it reached a critical condition.
FAA AD 87-21-08 should have mandated tnspect ten of a11. lap jOint$ per Boeing ASB 737-53AI039,Revision 3, instead of limiting the inspection of only the 1ap jOints at 5-4.

12.

13.

14. It was not determined whether Aloha Airl ines actually performed the required eddy current inspection ineolllpliance with AD87-21-08 or whether it was performed ineffectively.
IS. A properly

conducted eddy current inspection, p'erformed in accordance withAD-87-21-08 in November 1987, would have detectedadditio'1a' fatigue cracks in the holes of the upper rivet row of the lap jOint along S-4R.
fAA licensed

16.

Aircraft and Powerplantmechanics are not required to be knowledgeable tnthe maintenance and inspection of modern contemporary airplanes becau$e the training curriculum has not kept pace with aviation industry technology. There are human factors issues associated with visual and nondestructive inspection which can df!grade inspector performance to the extent that theoretically detectable damage is overlooked. Aloha Airlines management failed to rel;ognize the human performance factors of inspection and to fully motivate and focus their inspector force toward the crtlicalnature of lap joint inspection. corrosion control and crack detection. However, reports of fleet-wide cracks r'ece1ved by the FAA after the Aloha Airlines accident indicat!! that a similar lack of critical attention to lap jOint inspection and fatigue crack detection was an industry-wide def'jciency.

17.

18.

73

19. Because of the 'inexact instructions in AD 87-21-08, Aloha Airlines maintenance personnel did not replace the S-4R lap joint upper row CQuntersunk rivets with protruding head rivets.
20.

The NASIP inspection of Aloha Airlines in December 1987 did not reflect accurately the airworthiness of the operating fleet because the team failed to inspect adequately and report the physical condition of the fleet.

21. The PMI assigned to Aloha Airlines, although motivated toward his FAA surveillance tasks, was overburdened with other FSDO responsibilities and not suitably informed about the age and condition of the Aloha fleet or the nature of the Aloha operations. He lIIas therefore unable to provide sufficient impetus to effect necessary timely improvements in the Aloha Airlines maintenance program. 22. The PHI was not specifically trained to deal with the lap joint corrosion and disbonding problems of the 8-737. His efficiency was further eroded when he was excluded from the informational loop regarding Boeing aging aircraft inspections of Aloha airplanes and not apprised of the program between the FAA Aircraft Certification Service and Boeing regarding structural testing of an in-service airplane. Principal Inspectors have difficulty initiating safety improvements in air carrier operations and eaintenance programs outside the -regulatory compliance- approach, and they must resort to salesmanship and persuasion unless an enforcement is clearly viable. The Boelng SSIP did not include supplemental inspections on the fuselage mini.um gage skin due to damage tolerant design which was claSSified as ·damage obvious· or -malfunction evident- because of the controlled decompression scenario.

23.

24.

25. The aviation industry preaise that airplanes can be operated in a safe airworthiness condition indefinitely is sound only if operators hav& an effective inspection, corrosion control, and damage repair program.
26.

The current FAA and industry activities to address the aging airplane issue IlUst be continued to prevent accidents caused by structural failure.

3.2
_intenlnee

Probabla cause
progr_

The National Transportation Safety tioard deter.tnes that the probable cause of this accident was the failure of the Aloha Airlines to detect
the presence of significant

fatigue da-age which ultimately led to failure of the lap Joint at S-lOl and

disbanding and

74

the separation of the fuselage upper lobe. Contributing to the accident were the failure of Aloha Airlines management to supervise properly its .. intenance force; the failure of the FAA to evaluate properly the Aloha Airlines maintenance program and to assess the airline's inspecticn and quality control deficiencies; the failure of the FAA to require Airworthiness Directive 87-21-08 inspection of all the lap jOints proposed by Boeing Alert Service Bulletin 5B 737-53A1039; and, the lack of a complete terminating action (neither generated by Boeing nor required by the FAA) after the discovery of early production difficulties in the 8-737 cold bond lap joint which resulted in low bond durability, corrosion, and pramature fatigue cracking.
4. RECOMMENDATIONS

As a result of its investigation of this accident, the National Transportation Safety Board made the following safety recommendations: --to the Federal Aviation Administration: Provide specific guidance and proper engineering support to Principal Maintenance Inspectors to eva 1uate modifications of airline maintenance programs and operations specifications which propose segmenting major maintence inspections. (Class II, Priority Action) (A-89-53) !dentify operators whose airplane use differs significantly from the f1ight~cycle versus flight time relationship upon whicn the Maintenance Planning Docu.nent was predicated, and verify that their maintenance programs provide timely detection of both cycle and time related deficiencies. (Class II. Priority Action) (A-89-54) Revise the regulations governing the certification of aviation maintenance technician school s and the 11cens ing of airframe and powerp1ant mec;'anics to require that the curriculum and testing requirements include modern aviation industry technology.

(Class 11, Priority Action) (A-89-55)

Require formal certification and recurrent training of aviat ion maintenance inspectors performing nonde.struct ive inspection functions. Formal training should include apprenticeship and periodic skill demonstration. (Class II, Priority Action)
(A..S9-56)

Require operators to provide specific training: programs for maintenance and inspection personnel about the conditions under which visual inspections lIust be conducted. Requi re operators to

75

periodically test personnel on their ability to detect the defined defects. (Class II, Priority Action) (A-89·57) Develop a continuing inspec.tion program for those 8-737 airplanes that have incorporated lap joint te~inating ~ction (protruding head solid fasteners installed in the upper row of all lap splices) to detect any fatigue cracking that may develop in the Middle or lower rows of fuselage lap jOint fastener holes (for both the inner and outer skin panels) or in the adjacent tear strap fastener holes, and define the types of inspections, inspection intervals, and corrective actions needed tr~r continuing airworthiness. (Class II, Priority Action) (A-89-58) Develop a model program for a comprehensive corrosion control program to be included in each operator's approved maintenance program. (Class II, Priority Action) (A-89-59) Issue an Airworthiness Directive for 8-737 airplanes equipped with carbon steel engine control cables to periodically inspect the cables for evidence of corrosion and if there is such evidence, to accomplish the actions set forth in Boeing Service letter 737-Sl-76-2-A. (Class II, Priority Action) (A-89-60) Require that air carrier maintenance departments use the engi neering services available from the manufactuTPr or other sources to periodically evaluate their maintenance practices including structural repair. compliance with airworthiness directives and service bulletins, performance of inspect ion and qua 1ity assurance sect ions and overall effectiveness of continuing airworthiness programs. (Class II, Priorfty Action) (A-89-61) Revise the National Aviation Safety Inspection Program objectives to require that inspectors evaluate not only the paperwork trail, but also t~e actual condition of the fleat airplanes undergoing maintenance and on the operational ramp. (Class II, Priority Action) (A-89-62) Require Nat10nal Aviat10n Safety Inspection Program teams to indicate related systemic deficiencies within an operators maintenance activity when less than satisfactory fleet condition is identified. (Class II, Priority Action) (A-89-63)

76

Evaluate the quality ofF.AA surveillance provided by theprindpalinspectors as part of the National Aviation Safety Inspection Program. (Class lIt Priority Action) (A·S9-64) Integrate the National Aviation Safety Inspection Program team. leader ttl the closeout of the team findings. (Class U,Pri,ority Action)· (A-89-6S) Enhance the stature and performance of the priflcipal inspectors through; (1) formal management training .andgu 1dance, (2) greater encouragement and.•. backi"9 by headquarters of efforts by principal inspectors to secure the implementation. by carriers of levels of safety above the regulatory minimums, (3)1mproved accountability for the quality of the surveillance and (4) additional headquarters assistance in standardizing surveil1anceactivitie::;. (Class n, Priority Action) (A·89·66) Requ i re that all turbojet transport category afrplanes certificated in the future, receive full scale structural fatigue testing to a minimumof two times the projected economic service life. Also requ1rethat all currently certiftcatedturb(l;jet tra.nsport category airplanes that have not been fatigue tested to two lifetimes. be subject~ to such testing. As a result of this testing and subsequent ; nspect ion and analysi s, require manufacturers to identify structure susceptible to multjplesite damage and adopt inspectionprogl"ams appropriate for the detection of such damage. (Class II, Priority Action) (A-89-67) Discontinue classification of fuselage skin as "malfunction evident" or "damageobv1ous" on supplemental structural inspection documents. In addition, review all the remaining structurally significant items in the damage obvious category for possible inclusion in the Supplementary Inspectfon Program. (Class Il, Prior1tyAction) (A-89-68) Issue an Air Carrier Operations Bulletin for all air carrier flight training departments to review the accident scenariO and reiterate the need to assess airplane airworthiness as stated in the operators .null before taking action that may cause further damage or breakup of a damaged airframe. (Class II, Priority Action) (A-S9-69)

76

Evaluate the quality of FAA $urveilla~ce provfded by the principal inspectors as part of the National Aviation Safety Inspection Program. (Class II, Priority AcUon) (A-89-64)
Integrate the National Aviation Safety Inspection Program team ieader in the closeout of the team findings. (Class II, Pri.ority Action) (A-89-6S) Enhance the stature and performance of the principal 1nsp«.tors through; (1) formal management training and guidance, (2) greater encouragement and backing by headquarters of efforts by principal inspectors to secure the implementation by carriers of levels of safety above the regulatory minimums, (3) improved accountabil ity for the quality of the surveill ance and (4) additional headquarters. assistance in standardizing surveillance activities. (Class II, Priority Action) (A-aS-56) Require that all turbojet transport category airplanes certificated in the future, receive full scale structural fatigue testing to a minimumof two times the projected economic service 11fe. Also require that all currently certificated turb~jet transport category airplanes that have not been fatigue tested to two 1ifetimes, be subjectM to such testing. As a result of this testing and subsequent inspection and analysis, require manufacturers to identify structure susceptible to multiple site damage and adopt inspection programs appropriate for the detection of such damage. (Class II, Priority Action) (A-89-67) Discontinue classification of fuselage skin as "mal function evident" Dr "damage obvious" on supplemental structural inspection documents. In addition, review all the remaining structurally significant items in the damage obvious category for possible inclusion in the Supplementary Inspection Program. (Class II, Priority Action) (A-89-68) Atr Carrier Operations Bulletin for all air flight training departments to review the scenario and r'eiterate the need to assess airworthiness as stated in the operators .. nUll before taking action that may cause further damage or breakup of a damaged airframe. (Class II, Priority Action) (A~89·69) Issue an carrier accident airplane

77

--to Aloha AirHnes: Revise th'e maintenance program to recognize the -. high-time high cy'~les nature of the fleet operations and initiate mair;t@:i'lancenspection and overhaul i concepts based en realistic and acceptable calendar and flight cycle intervals. (Class II, t?riority Action) .(A-89-70) Initiate a corrositmprevention and control progf'am designed to affo~d . maximum protectionfrOllthe effeetsof .harsh operating environments (as defined bytheai rp 1ane manufacturer). (Class II , Priority AcUon) (A-89-71)
Revise and upgrade the technical d1vision manpower and organ1zat'ion to provide the necessary management, qua li ty assur-ance, engineeri"9, technical training and production personnel to maintain a high level ofairworth1ness of the fleet.' (Class II, Priority Action) (A-S9-72)

to. the Air Transport ',ssociation: Assist memberair carriers to establ hhmaintenance department engineering services to evaluate maintenance practices including structural repair, compliance with airworthiness directives and service bulletins, performance of inspection and quality assurance sections, and overall effectiveness of continuing airworthiness programs. (Class 1I" Priority Action) (A-S9-73)
BY THE IMTIOHALTRANSPOftTATIONSAfm BOARD Act11'9 Chail'llln lllllbar

Is/~

/5/ ~urnett lsI lsi John K.Lauber

.... ber

Joseph T~ ..... r
IIfIIIber

Na]]

/sl

Lemoine y, Dfckfnson. Jr.

78

statement:

Joseph T. Nall t Member, filed the following concurring/dissenting

While I concur with most of the majority's findings, 1 disagree with the probablfj cause and certain conclusions. Industry's best engineers

reviewed the carrier's records, knew of its high-cycle operations, and even inspected some of Aloha's 737 fleet. No one--not Boeing, Aloha nor the fAA principal maintenance inspectors--recognized or predicted the critical nature of the multi-site cracking or that the aircraft hull was about to rupture. If a Nfailure- occurred, it was a system failure. Could those who designed,fnspected or maintained the aircraft, given their knowledge at the ti_ of the accident,have reasonably foreseen this accident was about to hlppen? I think not. I would have preferred to cite simply -the presence of undetected disbanding and fatigue cracking" as the probable cause. I agree with the majority that contributing to the failure to detect the hull defects were systems, prcgrams or decisions of all the participants. But I emphasize these are simply contributing factors, not the probable cause of the accident. The majority's probable cause ;s too narrow and 1 therefore cannot agree that Aloha's maintenance program was the probable cause of the accident. I would have supported the following probable cause: The National Transportation Safety Board determines that the probable cause of this accident was the presence of undetected disbonding and fatigue tracking which led to the failure of the fuselage lap joint at S-lOl. Contributing to the accident were: the faflureof Aloha Airlines .. nag .. nt to supervise its maintenance force properly; the failure of the federal Aviation Administration to assess the quality and effectiveness of the Aloha Airlines maintenance program; the failure of FAA Airworthiness Directive 87-21-08 to require inspection of all the lap joints as proposed by Boeing Alert Service Bulletin 137-53AI039; and the lack of a complete tel"1linating action (neither generated by Boeing nor required by the FAA) after the discovery of difficulties in the early production 8-737 cold bond lap joint.

• .

lsI ~s.ph T Nall

..... r
e

-'14,1.9

79 APPENDIXES APPENDIX A INVESTIGATION NIl .HEARING

1.

Investigation

The Washington Headquarters of the Kational Transportation safety Board was notified of the Aloha Airlines accident within a short tiE after the occurreftCe~ A fun investigation team departed Washington, De at 0800 eastern dayl igbt time the following IlOrning and arrived in Maui 1400 Hawaiian stbdard tiE on the same day. The team was COI;'osed of the fonowing

investigative Metallurgy,.

groups: Operations, Structures/Systems. Maintenance Records, and SUT¥i7alFa~tors. In addition, specialist reports were prepared to summarize findings relevant to the CVRand FOR.

Atteftdants.
2.
Oft

Parties to the field investigation were Aloha Airlines. the fAA, the State of Hawaii Department of Transportation Airports Division, the Boeing COll8lercial Airplane COlllPany~the Air Line Pilots Association, the intemational Association of Machinists" and the Association of flight fublte Hnring

A 4-day public bearing was held in Seattle. Washington. beginning Ju1y 12, 19884 P.rti~$ representr:d at the hearing were the FAA, Aloha Airlines, 8oeingea-ercial Airplane Company. and the Air Line Pilots Association •

80

APPENDIX B
PERSONNEL INFORMATION captain
OIl

Robert

Lawrence Scnomstheiller

Robert t. Schornstheimer, 44, was nired by Aloha Airlines The captain holds Airline Transport Pilot (ATP) Certificate 110. 1958730 with airplane multier.gine land ratings and commercial privileges in airplane single engine land. The captain is type rated in the Boeing 737. The captain was issued a First Class Medical Certificate on November 2Si
May

31. 1971.

captain

1987. -nth no li.itations.

Oft

check in

8-737 silDUlator" at another major airlines facility. The captain a total of 8.500 hours, 6,700 hours of which were in the Boeing n7", During the last 90 days. 60 days, and 30 days before the accident, the captain bad flown: 107 hours, 4 mint.'1.es; 12 hours, 3 minutes; and 41 heurs, 23 lrinutes. respectively. In the 24 hours previous to the accident, the captatn bad flown 4 hours and 26 minutes a.nd had a total duty time of 8 hours
had flown
i:

On January 7. 1988~ the captain cOllPleted recurrent ground training tbe Boeing 731. On February 17, 1988~ the captain received a proficiency

and 50 ~i nutes.

First

Officer

lladellne ~

Tomptins

First Officer Madeline L TompkihSt 37. was hired by Aloha Airlines on June 4, 1979. The first officer holds AlP Certificate 1907395~ with co.ereial privileges in airplanes single and multiengine land and glider aero tow only.. The first officer was issued an FAA first Class Medical Certificate trlth no restrictions on January S, 1988, with the limitation that -Holder shall wear corrective lenses while exercising the privilege of her ainAfl certificate.The first officer qualified as a Boeing 737 first officer on June 30. 1979. "leT _st recent recurrent ground training and proficiency check were both cOIIPleted on April 8, 1988. The first officer had flown a total of 8,000 hours, 3,500 hours of which were in the Boeing 137. During the last 90 days. 60 days. and 30 days before the aCCident. the first officer bad flown: 189 hours. 29 .inutes; 128 h:ours~ 21.inutes; and 58 hours, 46 lI1autes. respectively. In the 24 hoUl"S previous to the aCCident, the lint officer had flown 1 hour and 5 .inutes and had a total duty time of

3 hour.s attd 20 llinutes.

f'lttbt A'tttndant

Clarabelle Lanstng.

Flight Attendant Clarabelle lansing, sa, was eIIflloyed by Aloha Airlines on August 111' 1951 and she had cOIIPleted recurrent e_rgency training Oft .sept""" 29, 1981. She was assigned exit l ..l for takeoff and landing.

81

Fligt.t Attendant Michelle Honda

Flight Attendant Michelle Ho~da, 35, was employed by Aloha Airlines on July ly 1974 and she had completed recurrent emergency training on December 9. 1987. She was assigned exit R-3 for takeoff and landing.

Flight Attendant Jane Sato-Tomita

Ai rUnes training

landing.
Inspector

Flight Attendant Jane Sato-Tomita, 43, was employed by Aloha on December 1, 1969 and she had completed recurrent emergency on January 5, 1988. She was assigned exit R-l for takeoff and Wilfred Y. K. Soong Inspector Wilfred
Y. K.

Airlines maintenance supervisor. He holds FAA Airplane and Powerplant License 1687694 and has 22 years experience in aircraft maintenance. He was appointed to his present position a~ an inspector in September, 1987. Mr. Soong testif'ied that his inspector training was received through on-the-jab instruction. Mr~ Soong's Aloha Airlines Formal Training Record contains one entry for the period subsequent to his appointment as an inspector, an NOI workshop presented by Boeing for 2 hours. Senior lead Inspector Edward Matsumoto Senior lead Inspector Edward Matsumoto was employed by Aloha Airlines in January, 1960, as a mechanic. He has served as an Aloha Airlines inspector and foreman. "e holds FAA Airplane and Powerplant License 1450079 and bas more than 33 years experience in aircraft maintenance. Mr. Matsmaoto's Aloha Airlines Formal Training Record is silent from 1968 UfIi,l1 Septembf:r 1", 1987, when he attended an HDI workshop presented by
Boeing for 2 hours.

June, 1966 in the 1jne maintenance

Soong was employed

department.

He has served as an Aloha

by Aloha

Airlines in

82

APPENDIX C

BOEING SERVICE BULLETIN 737-53-1039. Rev. 2

R!.VI S ION
, ,-:....

.-_
I...

! ...... .;_

....

.I. 1 i-;/"t

JI

~.

- ...1

TRANSMITTAL SHEET

•
JOINT

Boeing Service Bulletin 737-53-1039 This sheet tranSftits REVISION 2 dated February 8, 1974 to Service Bulletin 737-53-1039,BODY SKIN LAP INSPECTION AND REPAIR.
Thi. revisio4 constitutes a complete reissue.

NO'1'EI

~is

revision i. iaau&d to add a specific corrosion and damage inspection program, expand the effectivity to include aiiplan •• sealed in production, delete ultrasonic void and feeler 9aqe 1nsp,c~iona, delete several
fatigue

ultrasonic thickness measurinq inatl:Ulllanta and incorporate outstandin9' statt:s change notice ••
on wbi.ch lap
S3-1 0 17 ,

Airplanes

Bulletin

joints
OF

were

sealed.

per

Servi~e

CORROSION PROTECTION, a~ an adhesive deterioration and corrosion preventiw measure require corrosion and fatigl:e damage 1nspection per this revision.
of

SEALING

COLD-BONDED STRUCTURE FOR

Airplanes
Structural

issues

corrosion
pages

which joints were refastened per previous this bulleUn or th~ methods detailed in Repair Manual, S3-l~~1 # Figure :;, require inspection only per this revision.
on
of

All

relocated infor.mation. REVISION BlSTORr

this

revision

c:ontain new,

revised

or

Ori~iD&l I.suer Revision 11 "-vision 21

July ", October

February a, 1974

11, 19i~

1972

BO£IIilG COHM£tcZAL COS'fC»IER SUPPORT

AIRPLANE

COMPAUY

83

APPENDIX C

BOEING
TIE ~IHG CoYPAMY COMIIf:IIC.IAL."PLANE P.o. DOX :1707 SlATT1..E •• ASHJIoIGT()H" 12l CROUP

I
I.

ATA

S f R V I CfJ! ~ !:T7!7_~_'.39 II
53

SYSTEM:
SUBJECT'

737
.tomr

July 19, 1912 REVISION 2: February a. 1974

DATE:
...

B<mr SJrnl LAP

DlSPECrIOIO

_AU

Planning Information
A. Effectivity

1•

Airplane. ."-~fected This ~hange is applicable only to the airpl~~es 1l.Uodbelcw.

MOGEL'

SERIES

MFG. SERIAL ~U.

~EGISTAY NO.

AF (Alit

KZ04U'Gl22

FAAN(U

131-241

1%01

AND 20126

"",SO.II ANO "",522111

p~

7)1-$)-1039

, of

2'

84 APPBIDIX C

IOEIHG SERVICE 8U~LfTIN NO. 737-53-1039

MODEL

salles

£.

MFG. SEUAL No.

REGlSU't

NO.

J..

AN

Pl716

(AUt

At.&ERI E J

737-21)6
1'9306 THAU 19309

7T-VEC ...... 1&6 THRU N+MGI

AP (Ala CALIFORNIA' PGZl1-K27+ nl-293 PGZ1SlPG276 737-293

P&277~78

737-293

19713 AMO 1~11+

20334 ANO 2031S

N+6SG8 ANO "467&. N468At ANO N469AC

AQ CSOUTtwEST AlRlltfES CO.l PG279 731-Z93 20336 'GTO! 131-297 20345

ItX071 73 7-2 91C 20346
M {MGatTlNASI

I
I

PKfOS-PH'lC" .PH'7 UU ... 42 Q

737-217

P'toonpyoos

131-Z87
7)1-287t

2OS21 ANDZOS37

20403 THr<,U ~04o.

20401 1."0 2M08

194Q.8 AIfO 1~09

t. V-JTD ANO

lV-JNO ANO

L y..",." T'-'U l V-JMI t.1-JTO

LV-JNE

au UUATH£H5J
P""OOlC'.JOOl
P'I'3&1.

Jl\.1OJ

Ut-205C

131-205 731-2&6

;:~5& 2~12

t.tt-sus

Ut..SU.. l.Sut

ANO

IN-SUP

CP CUlUDIMf

PG501-~S04
PG50S

PACIFJ"

P~1'PJ032
IUSU.

137-217 737-217 7)7-217

731-2&673T-ZCO

19"4 tHitU 193&1 19888 20196 .... 0 20197

CF-CP& THAU ('F-CP~ tF-CPU c.F-tPV ANO

v-c.",z

£1

P't..02

lJfTLI

2019S
20070 ARO lOOn 20012' AND 20013 20074 10341 tHllU lOJ6S

'FRCfITUtU 'L ,c.zUPGe22

"1214

PG6ZKK624 H.:S rtC6l"'~'O ~J.l nT1I ~)T

731-2(,0 137-2(.0 nT-HI 7n"'21. 'fU-214 n7-l11

,n-2'22

1036. l0.92:

,,.11
1.an

M73ru MT)73F 'MAU _'U'IlF fd311~

..nut: •..,
,,1').OF

fl731.

ANO ..7319f

"'17OF

N1)Il"
N7J.'3It

'tll-',...1039 l

85

APPENDIX C

50fING SERVICE BULLETIN NO. 137-53-103~

(;USTO'IER

tUSTOH~R ,

NO.

NOBEL "
SERIes

MFG. SEIUItL

"fO.

REGISTRY

"0.

( I fIIOIMI IIC PH10~PH7U·

737-2 A 8 731-2~8 131-248

20480

THRU 20485

VT-EAG THRU VT-EAL

IN (IRISH lHTERNATI

PG2S1tPG252 PG253tPGZS4
PGZ55

I"
VI

PY32t-PY323 PYOU CIAA .. J Pto(13UPH13t

..~Z4 AND 20221 AND 131-2~8 20223 13T-248OC20.218 T~U 2~98

19425 20222 2O~20

£I-AS'

EI-AS."~ EI-ASK

ANOe

£1-.tS8
I-ASG El-ASE
EP-lfl:G

El-ASC T~u
EP-JIlf A~O

73"1-286(
737-130 131-130 Tn-BO 137-136

731-286

20500

AfeD 20499

EP-tRH

CL Ul=THAHSA' PAOO1-PAQ().4

"&0Q9-PA021

PAOOs-P ... ooa PA~Z2 PY)41-p.,346

737-2)0«

20253 TMAu zone

20231

19013THR.U 190H 1.,.7 fHltU 1902Q 19021 TKt.U 19033 19194

O-UO. TI1II.UD-AU. 9-&8£" D-A88E THilU O-A8H~ 5A~l

O-UEF THRU D-A!e!

D-I.6EA THRU 0-"8£0

JIO r A Ia

pens

AACAGASCM J

'J37-Ze.z

1111H.~ITl"E PL8~)
JeT f"'!."-Atit i
MIl C1IQ1IUI4 UlJ

1M &. SHIPPllilG1 131-2.0 ZOl~ nl-2HS 'nT-ill 2~S.l &frIO 20521

20<054
Ut ... TC "'-0 M CF-MA,W IN-IIIlTn

~lU'PG.;l'"

I
I

PUi.Z PY301urt102

""':WK.~

717-Z~C

U7·,tUC

198047 aSCQ

19i#o.a
20,",,96

Z04S5 ""0

CF-1U.Q

tF- .... "!IIO s

...'to

tF-"AP J.... Ol J&"05

"-!ItA"

,. u.n

PI;!1"'JilG51.
JIoGS"'l-KSIO

"G.'7:K","~T.

1It1ltllOllJ ~S7U.~572

n7-l81 J].7-2.1

ZOZZ. M010127
2('2741. AltO

'fll-l.l
131"'281

20413 &"0'
lCM!49

n'-zn ~S.~''''~'2131"211 ""1)

20501

t0506.lItI)

''''u 20"'Z ~o;e'

lC414

20277

,JAl40I.11tQ .lI... !;)] .. ~

J"400& .. tto

,,'1407

.I"~i2 A~
JAM1"

...... 08 '"..U

~A'''ll

"'.... 11

131"')"'10'19

86

APPEmlX C

SOEING SERVICE BULLETIN _0. 737-53-1039

CUStOMeR: CUSTOfIeR

& HO.

MODel , Sflt(eS

"fG.

Sf-Ul NO.

fffGISTltY

NO ..

IWI
I
I

(M.A.S.A.) PA099

tZ1-130

19473

.. 15N& 5

HZ UtEw leAUNOI "'199

PG.oI-PGo\03
P~436

~704

1':.753 737-222 19929 THRU 137-219 137-21 .... 10156 731-291 203"

19931

11(-H&(; ZIt-fU.l(

THRU ZK-NAE

ZI(-NAJ

PI fPIEOHON1J '~301-P'30S ~30~P"07

137-201

PG3Ga-P'GllZ

K049

'&035 "G041&~2

nT-20l 717-201 731-222 737-222

737-22Z

19423 AfO[) 20211 20212 THau 2tl216 19073 1<;547 ANO It;S45 19555 19920

19418 THflU 1'9lt22

H!'34N THRU N738N N740H ANO H7"'1 ..

N143M TH.U N141S "'752" N749N AND N7S1N

1ft9OIt91J

~s 'PAtlflC SOUTHWEST)

PG432tPGto33 ~:J5

731-21 .... 19682 AND 731-214 '137-21 ....

PG437-~440 PG~l

nT-a.

zon5

20157 ntRU 20160 203U

IIt"ZPS

M3799S lNO

1010'S

M'9... p~ THRU N987PS "98aj)~

,. IEAStE~
P.l0"2

PI.,t C"'LU ...... P,,",11

737r-ZU

202ii9 203%0 20300

19921
AP-Q

CX-iw. 2C3n
(F-E~ CF-E:>C

~sapG65Z

PROVlkCIAl) nt-let
7}7-2£1

.uo

tF-EPR

PW (P&(IFI' weSTERN.

P~],4 filK701PH702 t PS?Ol

7n-us
In-27K 1')7-2." 1)7-2"

137-214

l'l~ &1IfO 20142 197 .. ,

1970. 19701 ."0 202zc) 10329 tMaU 20331

CF-, ....

tF-'WO .NO Cf-II"E

IS-SlL ,tIIO Ilf'lV

(F-PMt

SA '~OUT~

K:J'3 ~)54-~!l.

Kuup;nz

.'.IC••'131-2"

IS-sa ..

1$-S8111 Zs.-Sflll

as-see

87 APPDlDll C

I'" ...................
PCOO1~
PCOQ),~

MODEL , SERIES

MFC. SEPtAL NO.

19764 AND
I 19769

R~GISTR.YNO.
9V-IFD ANO 9Y-8FE "iii) 9v-eFF t\f-ft;ec 9v-eae

'COOS
Til fOETA'

731-112 731-112 Ul-1!2

LTr.

19770 AND 19m

AKG

19111

1
I

a)1'O~1

1tH703'.... 704

731-2e.1C.

737-251

ZG2~

ZOZil

Z05'U 19679 AJIIO

AJQ

CR-au u.-aAC
.. r371S .. ntll n73713 CF-UO
ff9QOIU

AND

CIt-&&6

TS 1.IUMiA' PUSlUtUJ2 ""703

I'Cil'DUI'G702

131-2:91 737-297

13t-159

20209 ZG.24Z

19610 20210

AND N13717 AND N73712

rz

"UIE.sl~ t.TO.) 737-249;: ""OO4P'tOOS

ZOZC5 &110

20206

AND

Cf-TAN

..... (uniTED)

I
.,. ..

~P$040

~3-~.

~"

KOOl~

PGOSo ~;osl-PG01S

131-222 H7-22Z fU-ZZ2 73'1-222 UJ-2ZZ nl-212

1'904+ fHltu l,on 19016 ~u 1~~f'

19549 TMIU l'SS.

19C)3411 l'Htlu 19042

THilU ft90')'W

19556
19432

"9006V TKltU,"01400 MG)6U f'-V N9~OU h9()UU T"-V "'04IU ~5OU

TMiOU 1'!l9S6

M9051U T~U
, ...

N9075U
H-SM~

nASI SAO JlAoUVJJ ~TI-J'C475 n7-2Al

20092

nau

lCMe

:'I)-SN,

..

JQOl_"'_Z03 ~2Ze 'SUl

lII'C.2:n-~l.lO

flilESTlMJ

1:n-241 "Bl-t.T "'1'-2 •., ."".ZU

Tn-lOU

1.,.. 'Il« U :.,;00 I_a t.. u l~ll

20125 NU7 .,...,

zeq'''"

,..501. SCl!1i ~505W lMRU ~5l~1f4~21.

JMSl)iII T.. ., "itS)!""

'''u .....
&t,Q M90a

"'801 hOOnnoo)

'.IE»~JoatEOJ

''fOOI.

7),7-21OC Ul-21OC

,~
~
I""

&lIIO KIM

Ullla
_!lOT

.~111

88

APP£MDlXC

BOEING SERVICE BUllEtiN NO. 737-5~l039

C.USTOMER , 'USTOMEk NO.

ZD Utt%TANN1Al

MODEL t
SERIES

MFG. SERIAL MO.

ReGISTRY

NO.

KnUPG33Z

P,J'OUP.1102
PGQ36

737-Zo.
737-204 731-222

In-z~

lWlt

19T~

".,fl0UP.l104

.V3'UPY!62

731-204t

20234 AN6 lOZIZ lND

1'901~

AND IYTIZ

.H~

19710

G.~¥Rl.~ 6-&V.M
G-"Vfl:1f At«) G-Avltb

~"417 C~'W~~AND &-lINt ZOll9 G-'XNA AND w-AX~'

G-AZN%

•• ~r

not

fte follovug t:Jrl$ Hl:'ri.Oia .bcve boMft· GDHM

JtEl'UaLlC:· and MALAYSIA-SDiGAPOltE

.~Ud

available. bulkt.in
openton 1Jlela4ed in the original iaaue of lon9U. affected and therefore b"CIII. tobeeffectiYitya
U.JIG M:RI..DE.

731-53-1039

"

89 APPENDIX C

BOBING

SEKV%CE

BULLETIN

NO. 737-33-103S

2.

Spar•• Affected NO Boe1nq supplied

.par•••

a...... n.

on

longitwlinal lap joint.s on affected 737 airplanes wer. unufactured by plac1nq • Z'OOII temperature curUl., epoxy Idh •• l". (IMS 5-10) betw .. n t.M overlapp4ld IUn. prior to flush riveting the akin. tog&ther. The purpose of the bond w.. to iJlprove fatigue Ufe. Seven oper.tors have reported instcne.s of deterioration of the adh •• ive used to bond the lap joint. re.ulting in ~oint delamination lind' corrosion on thirty airplaniltl. This condit.ion val found on .irplane. wiL~a. low •• 3000 flight-hours. !be deteriorated ar... varied from small isolated pocket. to area •• ix feet or more in length. In .oat instance. thele area. ceuld be positively identified only aftar corrosian cau.ed exterior .kin bulqes, cracks or ai••inq fastener head.. Airplanes operating in wart!'! MOist climates are the most susceptible to adhes~v~ deterioration.

of joint deterioration reported to date, of the joint. had dave loped fatique cracks. SOJIIeof the airplanes with lar9'. are.. of delamination, had accu.ulated in exee.s of ZO,OOO cabin pre.suriztion cycles. Prolonged operation wit~ large .reas of delaiainatlon will eventually result in fati9Ye cra;ki~g. ~he t1De at which fatigue cracks will occur ~.pends on the 4e9r .. of' delUlination and the pre.ence of '~~rro.ier.. 10:..';" affective bar~, especially in the upper portior. o! the joint, will delay cracking. It b eMphasized t.hA~ !;!.-;i~\le crack vulnerability exi!llt.s only where there are i .. =~e areas of delamieation.
In
110M

all instance.

Laboratory tests have shown that for the li9'htest 9&ge skins (.036 inch) the theoretical fatigue life of a lAP joint with a luq. c.a of delamination would.l:xl reduced. The ~ed fati9Ue ~f. of • light gag. delaminated joint is attributed tc the .harp edge at the base of the fastener head COUPterslnk. He.vier gag •• kins in a joint with large del .. inated are.. exhibit higher fatigue atrength because the edge at the i)••• of the countersink U DOt ... sharp. Lap joints with outer .kin 9ages .O~'" iach and greater have .ati.factory fatigue life even ir. A aa.pletaly dela.inatad condition. However. th.se jo~nts are still .ubje~ to corrosion cWaa9'••

137-53-1039

90

APPEIID'IXC

BOEING SERVICE BULLETIN NO. 737-53-1039

8'1blMlq\'leftt ~ to original i •• u. of t.W.s bulletin the boftdecl lap joints were Haled in production to prevent

.,l.tur. contact with the cured adhesiv.. Meant .. nice experiuce with both factory .eal.ed and operator .. al.d joint.iAdic~tea that ... ling haa not be.n successful in pr.ventingadhealve deterioration. ~.re is .aa.e evidence ~t .,bt1ll:. MY .nter the joint. uound. the external head of joint fastenen. Becau .. of thh, the effectivity i. expanded to include all airplane. with bonded lap joints. o..cription

c.

Tni. hulletin revis10D (Revision 2) is issued to describe .1niaum corroeion and fatigue damage tn.pection PX09ra= for maintaining the .tructural integrity Of bonded lap joint.. The c::orrollion in.apection portion of the program conSists of anexteznal ~sual inspection for .kin bd;.. , cracks ot' mi.. ing fastener head. indicating substantial corrosion, and an exbrnal LPS-3 application. The fnquency of corroslon inspection should b. bued on operator exper!encebut .hould not .exceed .1x JIIOnth intervals tor joint. in tne lower lobe of the body and one year interval. for joints in the upper lobe. The fatigue damage inspection cansist. of an annual external close visual inspection of the joint outer skin adjacent to the head of the upper row of joint fasteners for cracks. 'l'hb inspection 1. applicable only to joints in which the outer .kin gage is le•• than .056 inch and should commence upon accumulation of JO,OOO cabin pressurization cycles. AlthOtl9h fatigue crllck vulnerability exi!Ots only in joint areas .1th large delamination.. this inspeetion is reco.'1Wended.t all joint ue •• where the outer .kin 9a9'e i. l••s than .056 inch becau ••• ervice experience ha, reve.led that there is difficulty in dbcriminatinq bondeCl joint area. from del&minatedareas when no corro.ion i. present in the delamination.

the

of delam1natior.. reported on sealed lap jOints, t..'l• is expanded to inclW'!. airplanes with bonded lapioint. that were s.aled in production. Airplanes on which lap joints were sealed per Service Bulletin 53-1017,
• ffectiri.ty SEALIBC

Becau ..

PaoTEa'lOH, are considered the s.e as tho.. airplanes ... led in p%Oduction &Dd require corrosion &Dd fatigue 4aMaqe inspection. Airplane. on which joints were rofa.tened per previous issues of this bulletin or one of tJM, _thad. detailed in Structural JlepaiJ: Manual. Subject 53-30-1 FiqQre S. require corrosion inspection only.

OP

COlD-BONDED

BODY

STRUCTURE

FOR.

CORROSIOK

737-53-1039 8

91

APPEJIDIX C

8O£XKG SERVlCZ aULLEflN MO. 737-53-1039

'!'be

ri1:l1 .. U.taftOJ:Y n .. lU, 4Uf1ou1~y bee... of _fadem

ocAcemiav ~t:. .. ~1ated tha~ t:J-. cIIIpi:.h of "'rat:e

a.ltrUODie deleted becatinl

'WOid
r

al tboagh

ad

sc:.e operator_ baY. UMd 1:.ba _y operaton !lave exper1ecced

<

&eler

qA9.

iMpaction.

u.

naalU.

a.v.ral

iupect10D

Qluaaotdc

UaitaUou

... det:eaw-t D. MIp!O!al

ain,

u...
beniD

tbi.cJtM.. -... "ing oparat:or. haft "ported w .-are aorro.loa oould DOt in.U.... ta.
t..CauM

Del II1.WlIdentud1n9' ad bbqJ~.taUcm of

'!be ftSta1r ct..ed.Ht!

Duipat:ed

ec.puy.

_1M.rUg

-.pxe.entative

baa

~D

approved by the rM at 'Ibe Boeing

z.

II!DpOW!r
_I'OaiaaU

....

follows.

....

--=.

and

Cft¥

aiae for i.ad1vlclual tuu Airplane

tIOWn-'1'i.me

ez.v
OeraUGD Conosioa Iaspaotion
% %

(Men)

.1_

Man-Hours

SHoun~
20 20

_0

.. tit_ Iupection

.0

raUvue .1ID4/or
ColToaicm"pair •
P.

•
Da.ber ~ location cd

Depmdut
nquincl.

Oft

.b.
•

•
of repairs

JlaUdal It: i. ~

- kice ad Aftilabll1!)t 'that

1;be

in paragraph II.A. lie furDiahacl or '-brie.ted aparatort, .a..tiA, .tock or purchuecl 4i~ectly
DOt

part. ad _udal.
r

identified

iad1a8~

1aalude4 itt npport of

~...

AacorcIiD9ly

w..

price IID4 4eUv.ry J:tull.Uft.

frc. data b

fro.

737... 1039 $3...

92

APPEIDIX C

IOIIHG SERVICE 8~IN

DO. 731-53-103'

G.

7oolin7 - Price and Availability .,_cial ultrasonic or low fhqOlncy eddy"'CU1'reft teatin, eqW.PMnt, Vilu-~k iUUllat10n toola e4 UBilink a.ad Bulbed Chuzylodt iutallation toola aay b. ftMdeel dllpen41n9 Oft operator'l eoane of ac:tloD. If aac::h equis-ent is 4... i.!'ecl, it 18 re~ 1:bat it be tumiahecl froll operator' I e:dating iD"lltcry' or pur-"ecl tinctly fro. Inda.uy lougaa. Accorclin,ly, prica anel del1ftzy data ia not included in sapport of thb bulletin.

II. !!.i!ht

and. •• lance
!label: I

Change in •• i9ht and. ba~ &Ilea ..,ill ckJpend on aDd li.e of repain and &mOunt of ••• lat applied.

and/or LPS-3

type

1.

Exiltin, e. b. c.

Datal

I
2.

Repair Manual D6-15565, Subject 51-10-2, 53-30-1, 53-JO-2, 53-JO-3 and 53-30-4. Boeing Non-Destructive Titlt Manual, DoCUMnt D6-'1'O, Part 1 and '.rt • Structural Boeing S.rvice

Cold-Bonded

Structure For COrroSion 'rotection

of this

Bulletin

53-1017, Se.ling

of

Newor reviaed data supplied in .upport

bulletiDs

None
J. Public.tiona pabl1cation: Affected

The .edification d.scribed herein affect. the fol1owin9

Publication 737 Structural Kanual

Chapter and/or Section Repair

53-30

737-Sl-10lt 10