SA-None File No. A-0004

AIRCRAFT ACCIDENT REPORT JAPAN AIRLINES COMPANY, LTD.

CONVAIR 880, MODEL 22M, JA8028 GRANT COUNTY AIRPORT

MOSES LAKE, WASHINGTON JUNE 24,1969

Adopted: June 17, 1970 / Report Number: NTSB-AAR-70-11

1 . 1.2 1.1

1.3 1.4

1.6 1.5

1.7 1.8 1.9 1.10 1.ll 1.12

1.14

2 . 2.1 2.2

1.13

1.15

3 .

lL4TIOEXL TRANSPORTATION SAFEIT BOARD DEPARlMENT OF TRANS€QRTATION

WASHINGTON. D . C .

COMTAIR 880. MODEL 22M. JA-8028 JAPAN AIR LINES COMF'ANY. IIPD . MOSEWASHINGTON GRANT COUNTY AIReORT

JUNE 24. 1969

TABIX OF CONTENTS

EE synopsis . . . . . . . . . . . . . . . . . . 1 Probable Cause . . . . . . . . . . . . . . 1 Investigation . . . . . . . . . . . . . . . 2 History of Flight . . . . . . . . . . . . . 2

Damage to Aircraft 5 Injuries to Persons 4

Crew Information 5 Other Dama@;e 5

Aircraft Information . . . . . . . . . . . . 6 Meteorological Information . . . . . . . . . 6 Aids to Navigation . . . . . . . . . . . . . . 6 Aerodrome and Ground Facilities 7 Couuuunications 6

Flight Recorders . . . . . . . . . . . . . . . 7 Wreckage . . . . . . . . . . . . . . . . . . . 8 Fire . . . . . . . . . . . . . . . . . . . . 9 Survival Aspects . . . . . . . . . . . . . . 10 Tests and Research . . . . . . . . . . . . . 10 Analysis and Conclusions . . . . . . . . . . 15 Analysis . . . . . . . . . . . . . . . . . . . 15 Conclusions . . . . . . . . . . . . . . . . . 18 (a) Findings . . . . . . . . . . . . . . . 18 (b) Probable Cause . . . . . . . . . . . 19

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .

Recommendations . . . . . . . . . . . . . . . 20 Attachment

Appendices A . Crew Information B . Investigation and Hearing

F i l e No. A-0004

NATIONAL TRANSPORTATION SAFEIT BOARD DEPARTMENT OF TRANSPOFSATION

AIRCRAFT ACCIDENT KEWRl

Adopted: June 17, 1970

CONVAIR 880, MODEL 22M, JA-8028 JAPAN AIR LINES COMPANY, LTD.

MOSES LAEX, WASHINGTON GRANT COUNTY AIRPORT

24, 1969

SYNOPSIS

Ja an~.Air. ,Li&es trainin& _- __ f l i g h t 90 of June 24, 1969, crashed at 1603 &hile executing a takeoff on Runway 32R. O f the f i v e crew- members and t ra inees on board, three received f a t a l in jur ies and two sustained serious burn injur ies . The two survivors were the f l i g h t instructor and the f l i g h t engineer.

Shortly a f t e r l i f t - o f f , the f l i g h t instructorzeduced zower on No. 4 engine t o check the t ra inee’s “nergency procedures, and t h e air- craft began t o yaw t o the right. This yaw-continued t o increase un t i l , a t a point approximately 6,500 f e e t down the runway, the r igh t wing went down and the No. 4 engine pod made contact with the runway. In a severe s ides l ip t o t h e r ight , the a i r c r a f t s l i d off the runway and thence i n a

bursting in to flames before it came t o rest on an eas te r ly heading. northerly direct ion through 2,600 f e e t of rough terrain, breaking up and

Four crewmembers escaped from the burning aircraft through a break i n the fuselage, but of these four, only two survived. The body of one t ra inee was l a t e r found i n the burned cockpit area. Ground impact and f i r e a f t e r impact almost completely destroyed the airplane.

was the &laxed~.corrective action during a simulated critical-engine-out takeoff maneuver resul t ing i n an excessive sTdeslip from which full re- covery could not be effected.

The Safety Board determines tha t the probable cause of t h i s accident

.. .. . . ~ ~ ~ .. ~ ~~~. .

IJ A l l times herein are Pacific daylight, based on the 24-hour clock.

- 2 -

1. INVESTIGATION

1.1 History of Flight

Japan A i r Lines (JAL) t ra ining f l i g h t 90, June 24, 1969, was t o be the f i n a l stage of a U-hour, captain-upgrading t ra in ing f o r three

p i l o t experience i n the Convair 880. The f i n a l check of these t ra inees copilots, each of whom was type rated and had a minimum of 1 year 's co-

was t o be a JCAE (Japan Civil Aviation Bureau) A i r Transport Rating Check,

June 24 f o r the a i r c r a f t and was scheduled f o r 1530 t o 1930. scheduled June 26, 1969. The subject f l i g h t was t o be the third of

t ra ining f l i g h t of the day. A l l turnaround checks were completed and the The a i r c r a f t arrived on the ramp at 1501 a f t e r completing i t s second

aircraft was readied f o r the f i n a l f l i g h t of the day.

Company records show that the crewmembers for Flight 90 reported t o the t ra ining center a t 1430 and spent the intervening time i n preparation f o r the f l i gh t . The crew consisted of an instructor p i l o t captain, the f l i g h t engineer, and three p i l o t trainees. Company records showed that a l l crewmembers had an adequate r e s t period pr ior t o reporting f o r duty.

information about a sticking fuel quantity gauge and e r r a t i c indications According t o the surviving crewmembers, the a i r c r a f t log ref lected

on the copi lot ' s a t t i t ude indicator involving er rors of several degrees of p i tch and roll. The captain was unable t o r e c a l l the number of degrees involved. He s ta ted he observed no abnormal readings on the horizon indi- cator while taxiing.

e i the r the captain o r the f l i g h t engineer during the prestart ing, ground operation, or takeoff r o l l .

No items affect ing the airworthiness of the a i r c r a f t were noted by

The prestar t ing checklist included the requirement f o r a crewmember, deplaned f o r the purpose, t o observe physically the operation of the rudder-boost system and proper movement of the rudders. The ins t ruc tor captain and flight engineer both s ta ted this was accomplished, as were a l l other pref l ight and ground checklist requirements. No discrepancies were noted by e i ther crewmember.

The instructor captain was seated i n the copilot 's seat; and one of the p i l o t t ra inees was i n the captain's seat. The f l i g h t engineer was at the engineer's station, and another p i lo t t ra inee i n the observer's seat. The t h i r d p i l o t t ra inee 's posit ion could not be determined; however, he was expected t o occupy a seat i n the passenger cabin during takeoff and landing.

Aircraft performance computations prepared by the crew f o r the f l i g h t were l o s t i n the ground f i r e . Reconstructed computations based on exis t ing

- 3 -

conditions a re as follows: V I speed--125 knots, vR--130 knots, V2--145

and tha t V2 was 140 knots "plus something. " knots. 9 The captain's recollection was t ha t VR was 132 o r 133 knots

Recorded information from the FAA-operated control tower a t Grant County Airport showed JAL 90 requested information as t o the act ive run- way, temperature, and time at 1537. A t 1545, the flight reported ready t o start engines f o r an IFR departure on a loca l f l i gh t . Wind vas re- ported t o JAL 9 as 240' at 15 knots with gusts t o 20, alt imeter 29.66. A t 1548, a loca l clearance was issued t o the f l i g h t and, at 1555, JAL 9 advised it was taxiing. The loca l control recording showed JAL 90 an- nounced i ts readiness f o r takeoff on Runway 32R ( r igh t ) at 1601:25 and takeoff clearance was issued at 1601:30. wind was again reported from 2 b 0 at 15 knots.

s tar ted what appeared t o be a normal takeoff a t 1602. He sa id t h a t ap- proximately 5,000 f e e t down the runway and a t a height of approximately 50 feet above the runway, the a i r c r a f t yawed t o the r ight approximately

touch the ground. The aircraft continued i n a northeasterly direction, 30". A t approximately 1603, he saw the r igh t wing dip and appear t o

There was considerable dust and what appeared t o be smoke. The control ler descended, struck the ground, and appeared t o groundloop t o the right.

stated tha t approximately 5 t o 10 seconds after the aircraft crashed, it burst in to flames.

According t o one of the two tower control lers on duty, the a i r c r a f t

The other tower control ler s ta ted that he observed Japanair 90 make what appeared t o be a normal l i f t - o f f . He said tha t immediately after l i f t - o f f , the a i r c r a f t began t o y a w severely t o the r igh t at an a l t i t ude of not more than 50 fee t . Approximately 6,000 f e e t down the runway, the r ight wing dipped t o the surface, the a i r c r a f t groundlooped off the r ight side of the nmway, s l i d in a northerly direction, and burst i n to flames within 10 seconds after coming t o r e s t .

takeoff and crash. The two tower control lers were the only known eyewitnesses t o the

A ro l l ing takeoff was made (from Taxiway 1, approximately 500 f e e t from The instructor captain described the following sequence of events:

the threshold end of Runway 32R). Upon reaching VR, "rotation" was cal led out and the a i r c r a f t l i f t e d off normally. According t o the horizon inst ru- ment on h is side, the rotat ion angle was 12'. He s ta ted t h a t shor t ly after

power lever very slowly t o check the t ra inee p i l o t on a one-engine-out l i f t - o f f , with a speed somewhere between VR and V2, he retarded the No. 4

emergency procedure. A s No. 4 power was f u l l y reduced, he had the feel ing that the aircraft was leveling off, not climbing o r sinking, and that at the same time it was turning t o the right, wings level . The captain s ta ted

2J VI means cri t ical-engine- failure speed; VR means rotat ion speed, and V2 means takeoff safety speed.

- 4 -

that upon retarding No. 4 power, the t r a inee p i l o t applied rudder i n the correct d i rec t ion but t h a t h i s correct ion was "a l i t t l e delayed," "rather l ight" , o r "not proper. " He sa id t h a t he took over control of the a i r c r a f t when he saw the yaw continuing " ju s t about when we were going off t he runway" ( t he nose of the a i r c r a f t turned beyond t he side of the runway, and only the grass area of t he a i rpor t was i n view).

The ins t ruc to r captain, i n describing the ac t ions he took t o stop t he yaw, s tated, "I f e l t it strange t h a t though I was pushing the l e f t

aircraft d idn ' t follow so. A t first (I used) moderate rudder and when mdder t o s top yawing, the ac t ion has been ineffect ive at al l , and the

yaw d idn ' t s top I pushed rudder moderate t o heavy - about th ree quarters rudder t rave l . I couldn't use ful l rudder because I was too c lose t o t he ground and a f r a id of contacting the ground with t he l e f t wing as a r e s u l t of r ad ica l action." So, hoping t o maintain the a l t i t u d e he had already a t t a ined and s t a r t climbing, he advanced No. 4 engine power. This ac t ion stopped the nose from continuing i t s turn, but it was at t h i s time t h a t t he aircraft contacted t he ground. The captain s t a t ed t h a t he then aban- doned t he idea of t ry ing t o continue l i f t- of f and decided t o make an emergency landing. During t he ensuing ground shocks and rapid decelera- t ion, the captain retarded a l l power levers t o id le , pulled t he No. 4

No. 3 "T" handle pa r t way out. f i r e - p u l l "T" handle and, t o t he best of h i s recollect ion, pulled the

captain s t a t ed t h a t it was because of the ex i s t ing crosswind, and it was the first takeoff of the day ( f o r t h i s pa r t i cu l a r t r a i n ing f l i g h t ) . He a l so inferred he used t h i s technique t o af ford t he t ra inee a g rea te r

viously, on infrequent occasions, used a slow movement of t he power lever opportunity t o respond t o the s i tua t ion . He s t a t ed t h a t he had pre-

when simulating a one-engine-out maneuver. He was not sure whether or not he had done so previously with the t r a inee p i l o t involved.

To explain why he had reduced No. 4 engine power very slowly, the

Both the captain and f l i g h t engineer s t a t ed t h a t engine power and other aircraft instrument indicat ions were normal throughout t he f l i g h t .

After the a i r c r a f t came t o rest, f o u r of t h e occupants escaped from the burning wreckage through a break i n the fuselage. Of these, two

tra inee was l a t e r found i n the burned cockpit area. succumbed t o f i r e a f t e r ge t t ing outside t he a i r c r a f t . The body of one

The accident occurred at l a t i t ude 47"13'30"N., longitude 11y018'l0"w., during daylight hours.

1.2 In ju r i e s t o Persons

In ju r ies Crew Passengers

Fa t a l 1 ( t r a inee ) 0 Nonfatal 2 0 None 0 0

- Other

2 ( t r a inee s )

-

0

- .

- 5 -

1.3 Damage t o Aircraft

The aircraft was demolished by impact, by disintegration a s it s l i d through rough, rocky te r ra in , and by the ensuing ground f i r e .

1.4 Other Demage

of Taxiway 5 on which the main wreckage came t o r e s t was damaged by f i r e . Funway 32R was gouged, one runxay l i g h t was smashed, and the portion

1.5 Crew Information

positions. (For fur ther detai ls , see Appendix A , ) A l l crewmembers were properly cer t i f ica ted and qualified f o r t h e i r

l i ne Transport P i lo t Cer t i f ica te with type rat ings i n the Convair 880 and Douglas E-8, and an instructor-pilot ra t ing issued March 14, 1967. His and company records showed he had accumulated a t o t a l p i l o t time of 7,639 hours, wlth 600 hours experience as an instructor-pilot , of which

proficiency check i n the CV-880 reflected an outstanding performance. 100 hours were obtained instruct ing i n the Convair 880. H i s most recent

Captain Kazuhiko Suda, aged 37, held a currently effect ive JCAB A i r -

mately 12 hours of t ra ining i n simulators and the actual a i r c ra f t fo r '%I December 1968 and January 1969, Captain Suda received approxi-

qualification as a CV-880 instructor-pi lot . He acted i n t h i s capacity during January, February, and June 1.969. During March, April, and May 1969, except for several t r i p s as a Convair 880 copilot , he served as a E-8 yaptain and instructor.

was 29 years old and held a currently val id XAB Commercial P i lo t Cert i - P l lo t Trainee J y u n i c h i w a t a , who occupied the captain's seat ,

f ica te . Company records showed he had a t o t a l time of 2,773 hours. Of t h i s t o t a l , 827 hours were acquired i n t h e CV-880 as a copilot , 32 hours of which were in t ra in ing and obtaining a type rating. I n the last 90 days, he had accumulated 154 hours i n the cv-880, including 11 hours and 43 minutes of captain-upgrading training.

In describing Trainee Murata's s k i l l level, Captain Suda s ta ted tha t he graded the t ra inee with "B minus" on the intermediate check of June 19, 1969, which was about the halfway point i n the scheduled 12-hour

Mr. Murata had received 8.5 hours of training, had the comnent, "OK f o r t ra ining program. Captain Suda's report of t h i s check, given when

ATR check. Depending upon the demands f o r l e f t seat capabili ty, judgment

t ra ining would be required (programmed time, I 2 hours). Captain Suda said and planning is weak." The report projected an estimated 14 hours of

tha t Trainee Murdca tended t o scan the instruments slowly during his f l i g h t t ra in iag and tha t it was necessary t o remind him of a l t i tude , heading, and

that the

the "rat

goin of t

the rudd a i r c 3Taw ' rudd, grow of rt atta: stop1 the 2

donei

t ion, emeq

No. : f i r e -

the f capts

a l so

vious OPPr

when not h

other

1

the bi

t raint succul

5 during

1.2 1

1

F N N

- -

- 6 - .

airspeed. He judged Mr. Murata's ski l l a s " jus t average, or just below average." Captain Suda estimated that he had given Trainee Murata 20 or more one-engine-out procedures during h i s training.

JCAB Flight Engineer Cert i f icate . He had been a f l i g h t engineer instructor

which 1,114 were i n Cv-880 a i r c ra f t . since March 11, 1968, and had accumulated a t o t a l time of 2,948 hours, of

Flight Fngineer Nobuhiro Uematsu, aged 31, held a current ly effective

1.6 Aircraft Information

factured September .l3, 1962, and had accumulated a t o t a l time of 14,278 The a i r c r a f t , a Convair 880, Model 22M, serial No. 49, was manu-

hours. It was powered by four General ELectric Model CJ805-3B axial flow j e t engines, with t o t a l times which ranged between 8,914 and 11,326 hours.

lat ions, and the records ref lected tha t it had been maintained i n an air- The a i r c r a f t was cer t i f ica ted i n accordance with exis t ing regu-

worthy condition.

of gravity, at 27.6 percent MAC (mean aerodynamic chord). These were within the allowable limits of 193,000 pounds maximum gross weight, and the center of gravity range of 20.6 t o 32 percent MAC.

The takeoff gross weight was computed a t 162,450 pounds, the center

The h e 1 load was Conoco J e t 50, which is equivalent t o J@+ or M I L J-5624E. Records showed the aircraft was serviced t o 71,500 pounds t o t a l fuel, d is t r ibuted symmetrically i n both wings. Takeoff f u e l was calculated at 70,000 pounds.

1.7 Meteorological Information

the G r a n t County Airport nearest the time of the accident were made a t Surface weather observations recorded by control tower personnel a t

1555 and 1610. Both observations showed scattered clouds at 8,000 fee t ; thin, scattered cirrus; v i s i b i l i t y 65 statute miles; temperature 74" 8. , dew point 38" F., and alt imeter s e t t i ng 29.66. The 1555 observation showed the wind from 250' at 15 knots, and the 1610 observation showed the wind from 2800 at 10 knots.

The natural l i gh t condition was sunny.

1.8 Aids t o Navigation

Aids t o navigation were not involved.

1.9 Commnications

Communications were not a factor .

-

1

M

T 0

U a 11 1:

t k vf

Bu I C

l o i s

an i n

1.:

f r( i n dm bul

que gra

. min

act1 w i t :

tha-

and incl

at s occu

l -

- 7 -

t a 20 or

b effective

rhours, of instructor

b f 14,278 I mnu-

axial flow ., 326 hours.

regu- i n an air-

he center e were ght, and

0 pounds or

lrsonnel a t ! made a t

.re 74" F., {,OW fee t ;

ration showed :d the wind

1.10 Aerodrome and Ground F a c i l i t i e s

Grant County Airport i s located 5.5 miles north-northwest of Moses Lake, Washington, at an elevation of 1,186 f e e t m.s.1. Publicly owned, the a i rpo r t i s operated by the Port of Moses Lake Commission. There a r e three runways, the longest of which is lbL-32R. Runway 32R used by t he f l i g h t involved is 13,500 feet long, 300 f e e t wide, l eve l , and i s constructed of asphalt and concrete. There a r e s t ab i l i z ed over- runs on each end of t h i s runway, 1,000 f e e t long. High-intensity runway l ight ing is ins ta l l ed . Runway 32L is a 3,100-foot day/vFR runway, con- verted from taxiway use t o l i g h t a i r c r a f t operations. It is located a t the northwest end of Fiunway 32R and is p a r a l l e l t o it.

located on t he a i rpor t . Tower personnel approved by the U. S . Weather Bureau record l o c a l weather obsenrations, and complete weather service is supplied from the FAA Fl igh t Service Sta t ion at Ephrata, Washington, located about 11 miles west-northwest of the airport.

A control tower operated by the Federal Aviation Administration is

and the main taxiway from the operations apron t o Runway 32R were both i n excellent condition.

1.11 Flight Recorders

Ground f a c i l i t i e s were found t o be well maintained. Fiunway 32R

from the severely burned cockpit area. The recorder, which was located

damaged by f i r e , but was decipherable after extensive cleaning of t h e i n the be l ly of the a i r c r a f t behind the nose gear well, was severely

burned f o i l .

A United Data Control Fl ight recorder, Model FA-542, was recovered

Exmination of the f l i g h t record per t inent t o the takeoff and subse- quent accident showed t h a t a l l four parameters were functioning. A data graph re f lec t ing t he time period of 1 minute 40 seconds, beginning 1 minute before l i f t - o f f , was plotted.

Correlation of t he four t r aces showed t h a t takeoff was i n i t i a t e d with a ro l l i ng start following a lef t tu rn onto the runway heading. The actual takeoff r o l l was determined as nearly as possible, considering tha t a ro l l i ng takeoff was made, t o have begun at time 0:20 (sec. ).

The appearance of a l l t r a ce s showed that the takeoff nm was normal and accelerat ion was steady, as evidenced by a near ly constant r a t e of increase i n the airspeed t race .

at an airspeed of 131 KIAS (knots indicated airspeed), and l i f t - o f f occurred between times 0: 58 and 1:OO (60 sec. ).

Correlation of t h e t r a ce s showed t h a t ro ta t ion began a t time 0:55

- 8 -

of a turn t o the r ight , which i n the next 5.5 seconds resulted in a A t time 1:01.5 (61.5 sec. ), the heading t race showed the beginning

heading change from 318" magnetic t o 014", a change of 56" at a rate of

the airspeed trace showed a drop i n airspeed from 140 KTAS t o 34 KIAS. 10" per second. During the &-second period of time between 1:06 and l:lO,

It was determined that the magnitude of this decrease i n airspeed re- sulted from a loss of dynamic pressure at the p i t o t head due t o s idesl ip .

The a l t i t ude t race ref lected that the a i r c r a f t reached a maximum height of 125 f e e t above f i e l d elevation at time 1:08.5 (68.5 sec.). By t h i s time, the heading trace showed the turn t o the r ight had stopped and a turn t o the l e f t had begun, which, i n 2 seconds stopped again, and a generally s tab le heading of 345" magnetic m s then recorded. Shortly thereafter, at time 1:lO (70 sec. ), and a f t e r res torat ion of dynamic pressure t o the p i t o t heads, the airspeed t race se t t l ed on a value of 122 KIAS.

The a l t i tude t race showed a return t o f i e l d elevation at time l : l 3 (73 sec.) , at which time the airspeed was recorded at 116 KCAS, and the heading at 347" magnetic. It m s determined that ground impact occurred at t h i s time. A l l traces beyond time l : l 5 a re considered t o be meaningless.

No cockpit voice recorder ms' ins ta l led i n t h i s aircraft.

1.12 Wreckage

Gouges and scrape marks, s t a r t i ng at a point 6,504 f e e t from the

began scraping the runmy f o r a distance of 656 fee t . Tire skid marks, threshold of the runway, showed where the bottom of the No. 4 engine pod

which began short ly before the end of the engine scrape marks, continued

marks showed tha t when the nose gear touched down, the a i r c r a f t vas banked for another 278 f e e t before leaving the runway. Both engine and t i r e skid

at a 37" angle t o the horizontal and 18" t o i t s direct ion of t ravel . When the l e f t main landing gear made contact, the a i r c r a f t m s at a 42' angle t o the horizontal and at a s ides l ip angle of 22". A s it l e f t the mway, the a i r c r a f t was at an angle of about 46" t o the m w a y and at a s ides l ip angle of about 22".

aircraft continued t o s l ide approximately 2,600 f e e t on a magnetic heading Shortly a f t e r leaving the mway, the tail skid separated and the

of 351" before it came t o rest on Taxiway 5. Over approximately the last 1,000 f e e t of travel, scattered fragments of engine pods, main landing gear components, wing fragments, parts from the structure of the lower section of the airframe, and two separated engines were found.

the aircraft l e f t the runway and beyond the point i n the wreckage path Evidence of fire appeared approximately 1,700 f e e t north of where

where disintegration began.

Q 1

1 :

1 1

1

: 1

3

1 I 1

i

1 I 1

1

01

1 i

<

- 9 -

rain at a s l igh t ly lower l eve l than the surrounding ground, the fuselage When the a i r c r a f t reached Taxiway 5, which cuts through the ter-

s e p r a t e d a t the t r a i l i n g edge of the wings, and i t s f o m r d section came

bulkhead forward, the aircraft was completely destroyed by f i r e . Only t o r e s t on a heading of approximately 100' magnetic. From the pressure

the empennage, engines, and other components scattered along the wreckage path escaped complete destruction by f i r e .

Measurement of jackscrews showed tha t the t r a i l i n g edge f laps slat system and spoilers were at a se t t ing of 22' and the horizontal s t ab i l i ze r at 3.9 units--the proper positions f o r takeoff. A l l cockpit instrument readings and control set t ings were destroyed by f i r e .

The rudder-boost unit was removed from the empennage and sent t o a bench-test f a c i l i t y at San Leandro, California, f o r functional checking. Functional tests of the uni t showed tha t it was capable of normal oper- ation. A l l other aircraft systems were damaged by h p c t and f i r e t o the extent they could not be checked f o r preimpact operational capabili ty.

All engines were separated during the ground s l ide and breakup of the a i r c ra f t . The No. 3 engine was found 2,000 fee t from the runway, the No. 4 engine 2,300 fee t , and Nos. 1 and 2 engines were found i n the main wreckage. A l l engines were subjected t o varied degrees of roll, tumbling, and f i r e damage.

A disassembly examination of the engines, performed at the General Electric overhaul f a c i l i t y , Ontario, California, showed t h a t the No. 1

percent, and No. 4 more than 94 percent; a l l engines were decelerating. engine was at 9 percent r.p.m. o r less , Nos. 2 and 3 at approximately 90

The f u e l valves were determined t o be i n normal takeoff positions. No evidence of a preimpact malfunction o r f a i l u r e of the engines was found.

1.13 Fi re - Evidence of ground f ire was found approximately 1,700 f e e t north of

where the aircraft l e f t the m w a y and beyond the point where dis inte- gration of the aircraft began. Upon coming t o rest, the wings and the fuselage erupted i n flames. The fuselage (except f o r the empennage) and the wings were almost completely consumed by f i r e . I n addition, a brush f i r e occurred which spread over an area of approximately 300 acres before it could be extinguished.

Response t o the crash alarm sounded by the control tower was immediate

type 530A pumper, one type 0-10 crash truck, a tanker, and a pickup truck and crash/firefighting equipment reached the burning wreckage at 1607. One

This equipment was augmented by a tanker truck provided by the Boeing equipped with dry chemical were dispatched by the a i rpo r t f i r e s ta t ion.

Company. Foam was directed on the fuselage section of the aircraft i n

- 10 - an attempt t o extinguish fire i n the area where survivors might be

gallons of water were directed on the burning wreckage. Mutual a id located. Approximately 3,500 gallons of premixed foam and 2,500

from rural and c i t y f i r e departments was obtained t o combat the brush f i r e .

1.14 Survival Aspects

they could not see outside the aircraft a f t e r it stopped. Attempts t o leave the a i r c r a f t through the cockpit s l iding windows, the main cabin passenger door, and the galley service door were unsuccessful because a l l of these ex i t s were jammed and r e s i s t ed . a l1 e f fo r t s t o unlatch them. An opening at the rear of the remaining cabin section, resul t ing from

escape from the a i r c ra f t . Although the area i n the v i c in i ty of the separation of the a f t fuselage section, was discovered and u t i l i z e d t o

fuselage break was engulfed i n flames from the burning wing, Captain Suda

mately 6-1/2 f e e t into the f i r e , and then ran i n t h e i r selected directions. and Engineer Uematsu looked f o r an area of l ea s t f i r e , jumped down approxi-

Both successfully escaped from the f i r e area, but sustained serious burn injur ies , primarily t o t h e i r lower extremities. Two of the t ra inees were the first t o leave the a i r c ra f t . These two apparently went t o the r ight , near the area of the l e f t wing, and were not able t o escape the flames. The body of one was found approximately 75 f e e t from the fuselage t o the northwest, and the other, approximately 50 f e e t f a r the r northwest, i n the grass f i r e area. The body of the t ra inee who was seated i n the ob- server's seat was found near the observer's seat , i n the burned cockpit area. According t o the ins t ruc tor ' s statement, before leaving the air- craf t he found nobody remaining i n the a i r c ra f t .

The surviving crewmembers s ta ted tha t due t o smoke and flames,

The a i rpor t manager and a securi ty guard, the first persons t o arr ive at the wreckage, found both survivors together a short distance northwest of the empennage section. Both crewmembers were taken t o Samaritan Hospikal, Moses Lake, by emergency vehicle.

Because of destruction by fire, no evidence was available t o deter- mine the reason f o r the jammed exits encountered by the crew during t h e i r attempts t o escape. However, e m i n a t i o n of the available s t ructure of the a i r c r a f t showed t h a t the flexing and twisting forces t o which it had been subjected during the disintegrating ground s l ide would have readi ly caused the locking pins of these ex i t s t o jam.

1.15 Tests and Research

i s t i c s of the Convair 880, Model 22M, under critical-engine-out conditions, demonstration f l i g h t s were conducted, with Board investigators aboard, i n a similar make-model a i r c r a f t a t the Federal Aviation Administration Aero- nautical Center, Oklahoma City, Oklahoma, on July 7, 1969. The a i r c r a f t

I n order t o obtain first-hand knowledge of the handling character-

- 11 - and crew were provided by the FAA. Participating in the demonstration

Region Flight Test hgineering. The demonstration consisted of a series flights were the FAA Coordinator and a flight test pilot from FAA Western

of takeoffs simulating the critical engine inoperative, with such simu-

the first takeoff , the gross weight was 160,853 pounds, fairly approxi- lations and asymmetric thrust evaluations also performed at altitude. On

mating the load conditions of JA-8028 at the time of the accident. The

pounds. The field elevation was 1,284 feet. last takeoff of the series was performed at a gross weight of l37,oOO

Six takeoffs during which three engine cuts at Vl, two at V,, and

was retarded with a rapid movement of the power lever. To correct for one at VR-were made. In each simulation, the downwind outboard engine

the yaw produced by these power reductions, it was necessary in the first four instances to apply promptly full, or almost full, rudder travel. At the lighter gross weights, a V, engine cut at 139,000 and a V, engine cut at l37,OOO, three-quarters and slightly more than one-half rudder travel, respectively, were required. Although less rudder travel was necessary during the.lighter gross weight simulations, it was noted that the yaw developed more rapidly and corrections had to be applied more promptly. The use of aileron control to assist in maintaining straight flight was required in all cases.

At an altitude of 9,000 feet, a Board investigator took the controls to determine the mount of rudder and aileron necessary to compensate for

were used by the FAA crew to equate performance to takeoff at field ele- critical engine inoperative conditions. A Vmc speed and other conditions

vation. On three successive maneuvers, alternating with rapid reductions of power on the left-hand and right-hand engines, the Board's investigator found it necessary to apply promptly full rudder and some aileron to main- tain straight flight.

simulate the slow reduction described by the captain of the accident flight. On each of these maneuvers, the onset of yaw was gradual at first, giving rise to a feeling on the part of the investigator that no urgency was required in beginning correction and that light rudder pressure would be adequate. When corrective rudder and aileron were ap- plied, directional control was maintained or restored. In some instances where full rudder and some aileron were applied, the yaw was arrested and resulted in over-control and entry into the dutch r o l l condition, which is characteristic of any swept-wing aircraft.

On three additional maneuvers the power was reduced gradually to

Results of the above-described tests demonstrated that during critical-engine-out maneuvers, it was necessary to apply promptly cor- rective rudder supplemented with aileron control to maintain directional control of the aircraft. The tests at altitude simulating engine cuts demonstrated that with the use of full rudder supplemented by aileron, a developing yaw could be arrested and the aircraft made to yaw in the opposite direction.

- 12 -

f l i g h t test p i lo t t o determine the power increases required t o overcome While a t a l t i tude, additional t e s t s were made by the Western Region

drag resul t ing from one-half-ball and one-ball s ides l ip displacements of the turn and bank indicator.

Pressure Ratio) se t t ing of 1.85 was required. The a i r c ra f t was then put For the steady s t a t e zero s ides l ip condition, an E.P.R. (Engine

in to a one-half b a l l s ides l ip angle and thrus t was increased t o maintain airspeed and al t i tude. This required a power increase t o 1.95 E.P.R. A one-ball deflection required an increase t o 2.09 E.P.R. The data

Book 11, of Convair document zc-22-028 (Aircraft Cert i f icat ion Data), which recorded from these t e s t s were correlated with Chart 3.13 i n Volume 111,

CT-659-4D). It is a chart showing the re la t ion between engine t h r u s t and i s included i n the ce r t i f i ca t ion data f o r the Convair 880 (FAA Project

E. P.R. This chart shows t h a t variation of E. P.R. and thrus t is v i r tua l ly l inear , i .e . , t ha t th rus t f o r 1.90 E.P.R. is 4,400 pounds thrus t per engine and 2.05 E.P.R. resu l t s i n 5,060 pounds thrus t per engine. Thus a change of .O5 E.P.R. resu l t s i n a change of 220 pounds of thrust per engine, a t a given airspeed and al t i tude.

as follows:

Ball E.P.R. #Thrust/Engine #Thrust Required/Engine #Thrust Required

A graphic presentation of the resu l t s of the above t e s t s is shown

Over I n i t i a l For 4 Engines Over I n i t i a l

0 i420

/1,060 1,680

0

4,240

Thus it w i l l be noted tha t an additional th rus t of 4,240 pounds i s required t o maintain constant airspeed and a l t i t ude when the turn and bank indicator i s displaced one-ball width, which is i n excess of the thrus t produced by one engine with the b a l l centered. This t e s t demon- s t ra ted the high thrust penalty incurred resul t ing from additional drag due t o s idesl ip .

Computed data pertaining t o the takeoff performance of JA-8028 were studied i n an effor t t o determine whether actual performance was or was not normal. The information shown i n the appropriate JAL chart and the FAA Flight Test Engineering Chart gives the computed points fo r V l , VR, and V ~ F speeds. Though there are differences i n the two se t s of

used f o r comparison with the f l i g h t recorder information. The three s e t s computations, the differences a re not considered appreciable and were

of data are:

11 VmF means l i f t - o f f speed.

- 13 - Vl - I25 Knots VR - 130 Knots VmF - 140 Knots

JAL FAA

28.6 see. 3,495 f t . 30.0 sec. 3,805 ft. 32.7 sec. 4,470 ft. 29.5 sec. 3,640 f t . 31.2 sec. 4,025 f t . 32.8 sec. 4,487 f t .

Flight Recorder 35.0 sec. 130 knots 38.0 t o 40.0 sec., 136-139 knots

The f l i g h t recorder t races show differences of 3.8 t o 5.0 seconds at VR

these speeds. However, because of the ro l l ing takeoff, it Was impossible and 5 t o 7 seconds a t V ~ F -- somewhat longer periods of time t o reach

trace. For t h i s reason, it was considered that the f l i g h t recorder values t o determine the a i r c r a f t ' s precise s t a r t i ng point from the f l i g h t recorder

shown could not be compared with the computed values with any appreciable degree of accuracy.

Japan A i r Lines uses an approved derating system f o r power set t ings fo r the purpose of increasing engine l ife. The derating involves a take-

figures, and i s incorporated i n a l l computed power set t ings shown i n t h i s off E.P.R. power se t t ing .07 below the manufacturer's performance chart

report (except f o r the three-engine climb from l i f t - o f f t o 50 f e e t ) .

peculiar t o the Convair 880 22M during critical-engine-out takeoffs was requested of the manufacturer t o determine what, if any, character is t ics may have been causal or contributory t o the accident. Specifically, the manufacturer was asked t o show at what point i n a s ides l ip the rudder would no longer be effective. In reply, a Convair Division letter of September 23, 1969; included performance data re la t ing t o maximum rudder

maximum at ta ined speed of 140 KIAS as shown by the f l i g h t recorder, a deflection f o r s ides l ip angle vs. airspeed. The data showed t h a t a t the

s idesl ip angle of more than 7" would resu l t i n the beginning of rudder

forces and causes the rudder t o streamline with the v e r t i c a l s tab i l izer . "blowback," where t h e rudder hinge moment is Overcome by slipstream

Regarding this data, the manufacturer s ta ted that it was based on calcu- lated data not acquired by f l i g h t t e s t s , but considered t o be repre- sentative of the ac tua l a i r c r a f t response.

Information concerning the a i r c r a f t performance and p i l o t techniques

Forwarded wi th the same l e t t e r was a paper writ ten by Fl ight Captain A. P. Wilson, Engineering-Production Flight, Convair Division, en t i t l ed "A P i lo t Reviews Swept-Wing J e t Transport Takeoff." The paper was used i n the manufacturer's Convair 880 p i l o t t ra in ing program. An excerpt of Captain Wilson's paper is considered pertinent t o this accident, and is quoted as follows:

"Direction should be controlled by rudder and by l i f t i n g of the dead-

operative engine helps achieve a lower Vmc. A t c lose t o Vmc, full engine wing about 5 degrees. The 5 degrees bank away from the in-

t o maintain direction. The a i r c r a f t can be trimmed as desired, the rudder must be used; as speed increases, l e s s rudder will be required

minimum necessary being tha t desired.

1

- 14 - "In the main, rudder must be used t o maintain direction, keeping the

and if s l i p is allowed t o increase t o about 16 degrees, the added s l i p t o a maxlmum of 5 degrees. If s l i p i s used instead of rudder,

produce a ro l l ing moment tha t i s suff ic ient t o offset l a t e r a l con- lift on the lower wing and the decreased lift on the upper wing

t r o l ; at 15' s l i p condition the rudder, due t o increased a i r load, folds back t o a point where it i s streamlined with the f i n . A t ly t o 18" s l ip , the ve r t i ca l s t ab i l i ze r i s subject t o s tal l and direc- t iona l control is lo s t . With asymmetric power, a turning moment exis ts , the turning force as the rudder streamlines or the f i n stalls i s no longer held i n check and the a i r c r a f t turns in to the dead engine. Any yawing in to these s t a t i c limits wil l , proportionate t o t h e y a w rate, aggravate the condition. The forward moving wing,

generates more lift, the backward moving ( r e l a t ive ly speaking) wing due t o i t s higher speed, and increased effect ive angle of attack,

generates l e s s l i f t , and these differences (combined with the ctiffer- ences i n lift due t o the s l i p ) cause the a i r c r a f t t o r o l l rapidly. A t low al t i tudes, it may be impossible t o recover from such a con-

from t h i s degree of s l ip . By holding a maximum of 5 degrees s l i p , dit ion; therefore, it i s wise t o maintain a considerable margin away

such a margin is ensured. This trait i s character is t ic of sweot-wins a i r c ra f t , and it is essent ia l t o be careful not t o expose onel; self- t o such a hazard."

The f l i g h t operations t ra ining manuals of Japan A i r Lines and other Convair 880 operators reviewed during the investigation did not provide specific information regarding the l imi t s of rudder control labi l i ty , recommended maximum limits of s l i p angle allowable, and other information, i n the explanatory d e t a i l found i n Captain Wilson's paper. However, the approved Manufacturer's Fl ight Manual f o r t h e Convair 880, under the emergency procedures section relat ing t o the missed-approach procedures with three engines operating, s ta ted in par t "CAIFmON I T IS IMPERATIVE THAT LARGE SIDESLIP ANGLES (YAW ANGLES) ARE NOT ALLOWED TO DEVELOP. TO MINIMIZE THE PROBABILITY OF LARGE SIDESLIP ANGLES (YAW ANGLES) DEVELOPING

or a f t e r VI)" s t a t e s i n part: The Japan A i r Lines Flight Training Guide, under "Take O F F ( 3 Eng at

wings level. (Side s l i p must be checked by rudder. Never bank too "In case of eng. out, apply suff ic ient rudder and ai leron t o maintain

steeply). "

Elsewhere in the Training Guide it is stated, without explanation, t ha t the maximum bank s h a l l be limited t o 5" at V, speed.

- 15 - 2. ANALYSIS AND CONCLUSIONS

2.1 Analysis

Invest igat ion of the physical evidence and records disclosed that the a i r c r a f t was airworthy, properly loaded, and capable of normal operation. The testimony of the surviving crewmembers showed tha t the f l i g h t progressed normally up t o t he point where the deter iora t ion i n performance due t o excessive s i de s l i p resul ted i n a loss of control .

Although the f l i g h t ins t ruc to r ' s evaluation of the t r a i nee was that his progress was s l i g h t l y below average and he tended t o scan t he ins t ru- ments slowly, the ins t ruc to r a l so s t a t ed that with the addi t ional in- s t ruct ion t o be given him on the f l i g h t involved, he expected that the t ra inee would successfully complete t he t r a in ing program. As pa r t of t h i s instruction, the ins t ruc to r intended t o check the t ra inee ' s per- formance i n a s M a t e d cri t ical- engine-out procedure. After this f l i g h t , the t r a inee was scheduled t o receive a JCAB A i r Transport Rating check.

Crew testimony revealed tha t t he takeoff proceeded normally u n t i l

The crew was properly ce r t i f i c a t ed and qual i f ied f o r the operation.

p i l o t retarded the No. 4 power l ever very slowly. He stated that the shor t ly d t e r l i f t - o f f when, somewhere between V, and V2, the ins t ruc to r-

t ra inee ' s applicat ion of rudder t o correct the resu l t ing yaw t o the right was "a l i t t l e delayed" and t h a t he assumed control from the t ra inee " ju s t about when we were going o f f the runway" ( a i r c r a f t heading such tha t the runway was no longer i n view). The f l i g h t ins t ruc to r stated that at first he used moderate rudder and when the yaw did not stop, he pushed about three-quarters rudder t ravel . He stated, "I couldn't use full rudder because I was too c lose t o the ground and afraid of contact ing the ground with the l e f t wing as a r e su l t of r ad ica l action." The above statement shows that the t r a inee ' s correct ive ac t ion was delayed and inadequate, and fur ther , that the f l i g h t ins t ruc to r ' s assumption of con t ro l and his correct ive ac t ion were delayed and inadequate.

It is believed t h a t t he flight ins t ruc to r ' s fear of ground contact with the lef t wing, had he used more correct ive rudder, was due most l i ke ly t o the r e l a t i ve ly small angle of roll permissible near the ground. However, i n view of the rapidly developing y a w t o the r ight , with conse- quent l ikel ihood of ground contact, his fear of using full lef t rudder i s improper i n this s i tuat ion. When he found t h a t his rudder ac t ion did not

He s ta ted that this act ion stopped the yaw but ground impact occurred stop the yaw, t he f l i g h t ins t ruc to r reapplied power on the No. 4 engine.

shor t ly thereaf ter .

l a ted cri t ical- engine-out procedures, gained by t he Board's inves t igators during the course of the invest igation, showed that prompt appl ica t ion of full, o r almost full, rudder is required t o correct f o r yaw, and that a i leron control i n maintaining di rec t ion is a l so required.

Familiarization with the correct ive techniques t o be used i n simu-

- 16 - The f l i g h t recorder heading t race shows that short ly after l i f t - o f f ,

a turn t o the r ight began which, a f t e r 5.5 seconds, ended in a heading change of 56". It is evident that during t h i s period, excessive s ides l ip developed which exceeded the l imi t s of rudder control labi l i ty . In view of the excessive s idesl ip , rapid yaw ra te , and low al t i tude, it i s con- sidered highly unlikely tha t recovery of aircraft control was possible

minimal corrective rudder vas being applied before the yaw had progressed a t t h i s point i n time. The rapidi ty of the yaw r a t e c l ea r ly suggests

t o excessive proportions. A constant rate of the yaw during t h i s period fur ther supports t h i s .

Engine scrape marks and landing gear t i re marks on the runway re- vealed that the aircraft was on a heading of 37" t o the r igh t of the runway heading and 18" t,o the r igh t of i t s direct ion of t r a v e l a t i n i t i a l ground contact. A s the a i r c ra f t l e f t the m m y , these marks showed the a i r c r a f t heading vas about 4@ t o the r ight of the runway heading and about 22' t o the r igh t of the direct ion of t r a v e l of the aircraft.

Convair performance data relat ing t o the Convair 880M m a x i m u m rudder deflection f o r s ides l ip angle vs. airspeed show t h a t at 130 KIAS, maximum rudder deflection will control up t o a 10' sidesl ip , and at 140 KIAS, t h i s capabi l i ty is reduced t o about 7" sideslip. Considering the s ides l ip angle of between 18'and 22' disclosed by the engine and landing gear marks a t impact, the a i r c r a f t was at tha t time i n a s ides l ip which was well beyond the l imi t s of rudder control labi l i ty . According t o the manufacturer's mgineering Production Flight Captain A. P. Wilson, at a 17" t o 18" side- s l ip , the ve r t i ca l s t ab i l i ze r i s subject t o s ta l l and direct ional control

be impossible t o recover from such a condition, and he recommends a maximum is los t . Captain Wilson fur ther emphasizes tha t at low al t i tudes, it may

allowable s ides l ip of 5" .

The manufacturer's Flight Manual fo r the Convair 880 and the Japan A i r Lines Training Guide present adequate information with respect t o basic control techniques during three-engine-approach operations and three- engine takeoffs at o r a f t e r V1. However, these manuals and those of other

vide specif ic information regarding the limits of rudder control labi l i ty , operators reviewed during the investigation of t h i s accident did not pro-

mation i n the explanatory d e t a i l found i n Captain Wilson's paper. recommended maximum l imi t s of s ides l ip angle allowable, and other infor-

The increased engine power required as a resu l t of one-half and one- b a l l s ides l ip angles, as demonstrated by t e s t s accomplished i n Convair 880 aircraf t , showed t h a t a one-ball s ides l ip increased the drag by an amount approximately equal t o the thrus t of one engine. Thus it was determined that the s ides l ip condition which was allowed t o develop during the f l i g h t not only adversely affected rudder control labi l i ty , but a l so resulted i n a substant ia l loss of aircraft performance as well. The company's policy of a .O7 E.P.R. decrease i n power ( t o increase engine l i f e ) , was being com- pl ied with by the crew at the time of the accident. This derating of

- 17 - engine power, however, is not considered significant to the accident in view of the excessive sideslip and rapid rate of yaw which occurred just before impact.

In an effort to obtain greater accident prevention value from this accident, the Safety Board endeavored to determine the underlying in- fluences that may have resulted in the instructor pilot's obvious delay in taking corrective action before the aircraft reached the critical side- slip from which recovery could not be effected.

with the characteristics of the aircraft during the critical engine-out The evidence indicates that the instructor was thoroughly familiar

takeoff maneuver. He was not only experienced in the Convair 880, but he also had the reputation of being an excellent pilot and a good in- structor. In addition, he had been an instructor in the aircraft for

critical engine-out maneuvers. Also, he had passed satisfactorily several months and, in this assignment, had given trainee pilots many

regular checks on his own proficiency.

to the possible effect of the slow retardation of power on the timely recog- nition of the onset of the yaw moment. This consideration could have

to the instructor's delayed actions. The instructor knew he was reducing definite merit with respect to the trainee's response but not with regard

power and that he was doing it slowly. He therefore would be ready for a less decernible response of the aircraft, and, indeed, the slower reduction

power reduction. of power would afford him greater reaction time as compared to a more rapid

As a factor in the delayed corrective action, consideration was given

Consideration was also given to the tendency of an instructor to delay corrective action to give the trainee a greater opportunity to respond to a given situation. This trainee pilot had had some difficulty in his training and the instructor was probably concerned about the trainee's progress. This is borne out to some degree by the instructor's reserved evaluation of the trainee, but more so by the fact that it was necessary to give the trainee an estimated 20 critical-engine-out maneuvers in about I 2 hours of training. There is reason to believe the instructor was still concerned about the trainee's performance, and used the slow power reduction technique not to make the aircraft response more difficult to detect by the trainee, but to afford him greater opportunity to respond to the situation.

with an instructor's normal desire to assist the trainee reach his ob- jective, and to afford him maximum opportunity to respond to a given problem situation, may have caused the instructor to delay taking cor- rective action until a point was reached where full recovery could not be effected.

The Safety Board is of the opinion that the foregoing, in conjunction

2.2 Conclusions

(a) Findings

1.

2.

3.

4.

5.

6.

7.

a.

9.

The aircraft was properly loaded, properly cer t i f icated, airworthy, and capable of normal operation.

A l l f l i g h t crewmembers were ce r t i f i ca t ed and qualified fo r the f l i g h t involved i n accordance with exis t ing regulations.

A i r t r a f f i c control, communications, navigation aids, air- drome, and weather conditions were not involved.

The operation proceeded without incident u n t i l takeoff, when the No. 4 power lever was slowly retarded t o i d l e i n order t o simulate the l o s s of a c r i t i c a l engine a f t e r V,, and pr ior t o a t ta ining V2.

!Phe yawing moment resul t ing from cut t ing a c r i t i c a l engine was not controlled by the t ra inee and resulted i n an excessive

p i l o t t o regain control were too l a t e t o be effect ive enough s ides l ip and a high r a t e of yaw. Attempts by the instructor-

t o prevent the accident.

critical-engine-inoperative maneuver is the prompt appli- Proper corrective action required f o r yaw produced by the

cation of full, or nearly full, rudder at the onset of yaw. Aileron control is also required t o maintain direct ional control.

For t ra ining purposes, the slow reduction of power during a critical-engine-out maneuver may be an undersirable technique, because it may resu l t i n a delayed recognition of the onset of yaw o r overcontrol when it is recognized.

A heading change of 56" i n 5.5 seconds toward the simulated inoperative engine was recorded by ' the heading t race shor t ly

t h a t corrective action6 by the t ra inee p i l o t and the in- after l i f t - o f f . The rap id i ty of t h i s heading change shows

s t ruc tor were delayed and inadequate.

The pronounced depression i n the f l i g h t recorder airspeed t race between times 1:06 and l : l O , ref lected a substantial l o s s i n dynamic pressure recovered by the p i t o t head. This pressure loss indicated the existence of an excessive side- s l i p angle. Engine contact and t ire skid marks showed a s ides l ip angle of between 18" and 22".

- 19 - 10.

11.

12.

13

14.

15.

According t o data from the manufacturer, s ides l ip angles

the rudder re la t ive ly ineffective f o r direct ional control. i n excess of 1@ will stall the v e r t i c a l f i n and render

According t o the manufacturer, a s ides l ip angle i n excess of 5" should not be permitted during the cri t ical-engine- inoperative maneuver.

The Manufacturer's Convair 880 Flight Manual cautioned of the danger of permitting excessive y a w angles t o develop during critical-engine-out takeoffs. The Japan A i r Lines Flight Training Guide described control procedures f o r the maneuver.

been more def ini t ive and explanatory with respect t o the These manuals provided basic information but could have

f lec ted i n Captain A. P. Wilson's paper on the overa l l l imi t s of rudder cont ro l lab i l i ty and sideslip, as re-

sub ject .

The a i r c r a f t was destroyed by impact with obstacles i n the groundpath and f i r e , which broke out during. the ground slide.

J d n g of the aircraft exits and f i r e complicated and delayed crew egress from the a i rc raf t . The accident was survivable, except f o r the occurrence of fire.

(b) Probable Cause

The Safety Board determines t h a t the probable cause of t h i s accident was the delayed corrective action during a simulated c r i t i c a l -

f u l l recovery could not be effected. engine-art takeoff maneuver resul t ing i n an excessive sideslip from which

- 20 - 3- REC-TIONS (I

Based on this accident and others which have occurred under similar circumstances, the Safety Board recommends to the Administrator of the Federal Aviation Administration, that through his A i r Carrier Inspectors,

take the following actions: (1) re-emphasize to the pilot personnel the all operators of the Convair 880 and similar type aircraft be asked to

characteristics of these aircraft during critical-engine-out maneuvers;

aware of safe and proper critical-engine-out procedures, the limits of (2) assure that flight instructors, trainees, and line pilots are well

bilizer stall; and (3 ) caution all instructor personnel to emphasize that sideslip angles, rudder availability, and yaw limits for vertical sta-

they must be most careful to avoid any tendency to delay corrective actions too long during critical training maneuvers even though the purpose of the training flight is to check the actions of trainees who must have an op- portunity to respond properly.

In connection with recommendation number 2, the Safety Board con-

Production flight, an excellent example of information, the essence of siders the paper written by Captain A. P. Wilson, Convair Engineering,

which should be included in training manuals and curriculums by all operators of large swept-wing four-engine turbojet aircraft of the Convair 880 type.

In addition to the foregoing, as the result of the several above-

made other recommendations to the Administrator. These recommendations referenced accidents involving engine-out maneuvers, the Safety Board

and the Administrator's response thereto are included as Attachment 1 of this report. At the present time, deliberations on the recommendations are continuing.

BY THE NATIONAL TRANSPORTATION SAFETY BOARD:

Tune 17, 1970

JOHN H . REED C h a i r m a n

OSCAR M. LAUREL Member

FRANCIS H. McADAMS Member

LOUIS M. THAYER Member

Member ISABEL A. BURGESS

4 APFZNDIX A

Crew Information

A i r Transport Rating No. 313, issued September 5, 1960, and Medical Cert i f icate (Class I) No. 40335, va l id until October 31, 1969. He holds JCAB 3rd Class Aeronautical Radio Operator Cert i f icate No. 335, issued Januaq 10, 1956, with Medical Cer t i f ica te No. 31060, valid until October 31, 1969. Captain Suda was employed by Japan A i r Lines on April 1, 1954, f r o m the Civil Aviation College of Japan. His employ- ment record showed he had. been assigned as copilot on E-&, -6, -7, and Convair 880 aircraft, and t h e r e d t e r as captain on Convair 880 on A u g u s t 2, 1963, and E-8 captain January 10, 1966. He received h is Instructor P i lo t Rating on March 14, 1967. Company records show Captain Suda had accmu- la ted a t o t a l p i l o t time of 7,613 hours. O f t h i s to ta l , 2784:28 hours were as captain, 4245:43 were as copilot , and 583:17 as ins t ruc tor p i lo t , as of May 31, 1969. According t o Captain Suda, h i s personal records re- f l e c t an additional 25:49 hours of CV-880 instructor time, making a t o t a l

Captain Suda s ta ted he was assigned as a cv-8% instructor i n December 1968, of 100:14 hours as a Convair 880 instructor at the time of t h i s accident.

March, April, and May 1969, he was serving as a E-8 captain and instructor . and acted i n this capacity during January, February, and June 1969. During

Washington. Captain Suda had received the following 01-880 ins t ruc tor He w&6 then assigned t o the Convair 880 training program at Moses Lake,

t ra ining at Delta A i r Lines, Atlanta, Georgia --I2 hours; f l i g h t t ra in ing t ra ining in 'addi t ion t o l i n e p i l o t t ra ining and experience: Simulator

Moses Lake, January 7, 1969 --3 hours and 23 minutes. He received a p i l o t a t Nagoya, Japan, December 28, 1968 --2 hours and 40 minutes; and at

proficiency check on December 28, 1968, i n the Convair 880, which included engine-out takeoff. The check p i lo t ' s comment on t h i s f l i g h t stated: "In

t h e last cv-880 check r ide the proficiency shown by t h e checkee was out- sp i t e of the fact t h a t this check r ide was made after extended period since

standing." Company records show tha t during the 90 days pr ior t o the accident, Captain Suda acquired 110 hours p i l o t experience, of which ap- proximately 25 hours were i n the cv-880.

Instructor P i lo t Captain Kazuhiko (W) Suda, aged 37, holds JCAB

Cert i f icate No. 2150, issued June 8, 1967. xis Medical Ce r t i f i ca t e Pi lot Trainee J w i c h i Murata, aged 29, held XAB C o m m e r c i a l p i l o t

Class Aeronautical Radio Operator Cer t i f ica te No. 1954, issued October 25, (Class I) No. 30818 was val id u n t i l October 31, 1969. He held JCAB 3rd

1967, with Medical Cert i f icate No. 31260, va l id until October 31, 1969.. His JCAB Instrument Rating No. 937 was issued February 21, 1968. Mr. Murata was employed by Japan A i r Lines on October 1, 1966, from the Japan Self Defense A i r Force. Company records show he was assigned as a student p i l o t on April 1, 1967, and as a CV-880 copilot on February 22, 1968. These records show he had acquired 1763:06 p i l o t hours i n the J.S.D.A.F. and 1010:14 p i l o t hours with Japan A i r Lines. H i s t o t a l p i l o t time as of May 31, 1969, was 2773:20 hours. He was assigned t o captain-promotion t ra ining on June 12, 1969. Mr. Murata's records show he had 82~~1.7 co- p i lo t hours, includj.ng 32:22 fo r a cv-880 type rating. He had acquired

- 2 -

154:25 hours i n the CV-880 i n the previous 90-day.period. An in te r - mediate check on June 19, 1969, conducted by Captain Suda a t the halfway point i n t h i s training, showed Mr. Murata was progressing a t a r a t e of average t o s l i gh t ly below average and would require about 14 hours t o complete the course at the required s k i l l level. Programed time i s

Murata tended t o scan the instruments slowly, and it was necessary t o 12 hours. Instructor P i lo t Captain Suda's evaluation indicated Trainee

airspeed during maneuvers. I n addition, he tended t o ro ta te the aircraft remind him t o cross-check f o r overal l control of a l t i tude, heading, and

too rapidly f o r takeoff, although t h i s problem had been corrected on previous f l igh ts . The ins t ruc tor ' s comment on the intermediate check

b i l i t y , judgment and planning i s weak." follows: "O.K. f o r ATR check depending on the demands fo r l e f t seat capa-

Cert i f icate No. 173, dated November 7, 1963. His Medical Cer t i f ica te Flight Engineer Nobuhiro Uematsu, aged 31, holds JCAB Flight Engineer

No. 3233, was issued October 31, 1969. He also holds Firs t- class Aircraft Maintenance mgineer Cert i f icate No. 839, dated April 19, 1966. Mr. Uematsu was employed by Japan A i r Lines on April 1, 1956. During the course of h i s employment, he was rated as a Flight Engineer i n E-6, -8, and CV-880 aircraft, the l a t t e r ra t ing issued February 7, 1967. On March 11, 1968, he was issued a Flight Engineer Instructor Rating. Company records show tha t Mr. Uematsu had accumulated a t o t a l time of 2948:13 hours as of May 31, 1969, of which 1,114 hours were i n Cv-880 a i r c ra f t .

included i n the company's report. Crew information concerning Pi lo t Trainees Kawase and Yoshida is

4

Investigation and Hearing

1. Investigation

this accident shortly after its occurrence. An investigation team The National Transportation Safety Board received notification of

composed principally of personnel of the Field Investigation Division of the Safety Board's Bureau of Aviation Safety was immediately dis- patched to the accident scene. Working groups were established to

Systems, Structures, Maintenance Records, Human Factors, and Witnesses. investigate the accident in the areas of Operations, Powerplants,

Participants in the accident investigation were representatives of

Administration, Convair Mvision of General Dynamics, Inc., and General the Japan Civil Aviation Bureau, J a w Air Lines, the Federal Aviation

Electric, Inc.

The on-scene. phase of the accident investigation lasted approximately 10 days. The accident inquiry WBS conducted in accordance with the pro- visions of Annex 13 of ICAO (International Civil Aviation Organization).

2. Hearing

A public hearing was not held in connection with the investigation ! of the accident.

3. Preliminary Reports

accident. There were no preliminary reports issued in connection with the

Administrator Honorable John H. Shaffer

Federal Aviati.on Administration Department of Transportation Washington, D. C. m590 Dear Mr. Shaffer:

Since September 13, 1965, we have investigated f ive major alr car r ie r accidents which occurred while conducting simulated engine(s)- out t ra in ing close t o the ground. These accidents have resulted i n t o t a l destruction of t he aircraft; a l l but one resul ted i n loss of l ife, while four resulted i n serious damage and loss of l i f e on the ground.

t

A s pointed out i n your Notice of Proposed Rule.Making 69-14, issued March 28, 1969, other l e s s hazardous means of accomplishing t h i s type of t ra in ing are now available.

simulating v i r tua l ly every f l i gh t situation, including cont ro l lab i l i ty character is t ics of specific a i r c ra f t i n f l i gh t w i t h one o r more engines inoperative .

Present generation f l i gh t simulators provide the capabili ty f o r

may have t o be conducted under actual. f l i gh t conditions, despite t h e The Board believes tha t a very limited amount of proficiency t ra in ing

advanced simulator technology presently available. The degree of a i r c ra f t

maneuvers were t o be conducted at a safe a l t i t ude above the t e r r a in . vulnerabili ty and attendant hazards would be appreciably reduced if such

Should an actual f a i lu re of a c r i t i c a l system occur under these conditions, safe recovery could most l i ke ly be effected without ac tua l derogation of the trainj.ng value.

t o safety during t ra in ing f l i gh t s . The analysis of several t ra in ing accidents disclosed tha t an additional crevmember had participated i n the malfunction analysis and remxlia.1 actions taken by the crew. We do not ques thn the additional crew?!emnbcr’s qualifications in any of thcse

We a r e a l so concerned w i t h another factor which may be detri~mental

ATTACHaNT, Page 2

Honorable John H. Shaffer - 2 - December 18, 1969

situations, but we do question whether h i s part ic ipat ion appreciably contributed t o the crew's training.end efficiency. I n fact , one of our investigations indicates t ha t t he preserlce of t he additional crew- member may have -distracted the p i lo t at a c r i t i c a l time.

changes t o Federal Aviation Regulations Parts 61 and 121, as outlined i n Notice of Proposed Rule Making 67-14, t he Board recommends the following regulatory changes:

I n addition t o our comments forwarded t o you regarding proposed

1. A 2 1 maneuvers requiring simulated engine(s)-out operation ofr a i r c ra f t close t o the ground should be conducted i n and, t o t h e maximum extent possible, l imited t o f l i gh t simulator t ra in ing devices.

2. The Administrator should establ ish a minimum a l t i t ude above t h e t e r r a in for those simulated engine(s)-out maneuvers which must be performed i n f l i g h t .

3. Equally qualified crewmembers not d i rec t ly invo3.ved i n the t raining operation should be prohibitcd from the f l i gh t deck area on a l l proficiency f l i gh t checks.

with your FS-452 staff. The areas of concern have been discussed by personnel of our staff

Sincere1.y yours,

Original signed by #John H. Reed

John H. Reed Chairman

Y 0 OEC 1959 Kororable Jchr. H. Reed

Department of Transportation Chairman, Natiotnl Transportation Safety Board

\dashington. D. c. 20591

Dear Nr. Cimirnzn:

Thank you For yotm 1.etter of 18 December 1969, concerning a i r ca r r i e r accidents involving sin!nlated engine-out training.

Although the state-of- the-art o f a i r c r a f t simulation provides, as you say, the capability for simulating almost a l l t raining maneuvers, an adequate visual syntcm .is required to simulate manewers such a s engine fa i lu re on takcoff and the circl ing approach to landing. A very l iu i t ed nmtbcr of these simulators are i n use today.

X believe you vouLd agree tha t the regulatory history shows increasing training c r d i t : :or a i r c r a f t si~;lul.ators as the stat@-of-the-art has ndvauced. Ec~inning as carly as July 1955, the Civil Aeronautics A3ninioEraltion i n i t i a cnd action which led t o adoption of a rcguf.ation on 15 :.l:wch 1957, v'nich permitted the a i r carriers to subtjtitute an approved

checks. This rule was subsequrntly snrondcd so tha t only cer ta in lov simulator courw for one of tile required sculi-annual p i l o t proficicncy

a l t i t c d e s!itncuvcrs need be acccq>lislled i n an airplane during the a l t m a a t e pKofiCiCnCy check (e .%. , 'ILS approach, engine fa i lu re on takeoff, c i rc l ing approach), providcd tho balance of the rcqirircd nmncuvers mre cocducted i n an approved a i r c r a f t s i r d n t o r . I.Iore recently, regulations became effec- t ive on 15 Apr,iI 1967, which for the tirst time allowed cer ta in emergency maneuvers to be cndtlctcd i n a simulator for the a i r c r a f t type ra t ing check. On 1 Nay 1968, the rules were m!e~\ded t o chanee the in- f l ight requirements fo r turbojets re la t ive t o the approach and l a n d h ~ maneuver &€b the sictulated fa i lure of 50 percent of the wai lnb le por-crplants.

Simulating an ensine fa i lu re during tak",off a t a l t i t u d e i s not a satis- factory subst i tu te fo r accoapliching' the rmneuver in an airplane during an actual takeoff or i n a visual sind.ator since, when done a t a l t i tude ,

amendments to Parts 61 and 121 of the Federal Aviation Regulations pe rmi t the realism of an actual engine fa i lu rc during takeoff is los t . Recent

t h i s xaneuver to be conducted during type ra t ing and proficiency checks,

1

- -_l_ - .-

ATTACHMENT, Page 4

L

i n a simulator with an approved visual system which should encourage the a i r car r iers to procure these systems. Since airplane performance is baqed

handle an engine-out s i tua t ion during regular operations. The s a w rat ionale on engine fa i lu re during takeoff, we must ensure tha t p i l o t s can safely

ments allow the person conducting the check the option of giving the p i l o t a lso applies to the engine-out missed approach. However, the recent mend-

being checked e i ther an all-engine missed approach o r an engine-out missed approach, which has the advantage of introducing the element of surprise when the l a t t e r maneuver i s given.

The amendments further provide for f l i g h t t r a i n i m and a f l i g h t check i n

of proficiency, as appropriate, before the trainee advances to the ac tual an airplane simulator to the p i l o t in cor~mmd o r second i n comaand level

airplane. Thus, the ru le contemplates a thorough knowledfie of each requ1re.d maneuver and procedure and a predeternined level of s k i l l prior t o p- “rfono- ing thcn i n actual f l igh t . Therefore, the performnce of t he naneuvers and procedures by the trainee on a one-time basis i n the airplane during i a i t i a l or t ransi t ion fl.iejht t rainin6 is, i n pract icc l e f fec t , a deternyination tha t the proper transfer of learnins from tho sj.mul.ator to the airpl.ane has heen achieved.

Although as you point out, one acci.dcnt may have invo~ved . .d i s~mct ion of tile p i l o t by an additionat crcwniemb’er on the f l i g h t deck,’ therk’hnkct bcfn ot11er cases during t raining f l igh t s where an additiorlal crc?r,se:mL!cr has aler ted the

accomplished i n pairs so tha t each p i l o t mcy observe the other durine the;+ p i lo t to an unsafe condition. Flight traicin: of c+mmer:>bero i s usuc~l.ly

respective periods of f l i g h t instruction. .Such t raining i s highly desirn].,ie and edvantageous to the p i l o t s involved. Nevertheless, we will advise our Principal Opcrations Inspectors to see to i t tha t the i r assigncd a i r carr icro have isaucd adequate instructions t o ensure ‘that other creerambers on the ~~ ~~

during t ra in ins f l ights . f l i g h t dcck do not in ter fere with the performance of the operating crcmcmbers

Your int.erest in this olztter is appreciated.

Dl D . Thomas Acting Adrni.nisl:rntor

\