I

NEW ZEALAND

AIRCRAFT ACCIDENT
REPORT No. 1647

DOUGLAS DC-8-52 : ZK-NZB

Auckland International Airport
I

4 July 1966

ACCIDENTS 1NVESTIGA4TIONBRANCH
DEPARTMENT OF CIVIL AVIATION
WELLINGTON, NEW ZEALAND
 -

Accidents Investigation Branch,

Department of Civil Aviation,
Wellington.
The Minister of Civil Aviation.

SIR,
Attached hereto is the report on the investigation made into the
circumstances of an accident involving Air New Zealand Ltd. Douglas
DC-8-52 aircraft ZK-NZB at Auckland International Airport on 4 July
1966.
This report is submitted pursuant to regulation 11 (1) of the Civil
Aviation (Investigation of Accidents) Regulations 1953.
0. J. OBRIEN,
Chief Inspector of Accidents.
10 September 1966.
APPROVED. 4

JOHN MCALPINE,
Minister of Civil Aviation.
14 September 1966.
 ACCIDENTS INVESTIGATION BRANCH
DEPARTMENT OF CIVIL AVIATION
WELLINGTON, NEW ZEALAND

AIRCRAFT ACCIDENT
REPORT No. 1647

DOUGLAS DC-8-52: ZK-NZB

Auckland International Airport
4 July 1966

NOTIFICATION
1.Notification of the occurrence of the accident was received by
telephone from Air New Zealand Ltd., owners of the aircraft, at
1615 hours on 4 July 1966.

TIME, DATE, LOCALITY

2.The accident occurred at 1559 hours on 4 July 1966, the aircraft
making initial impact with the ground 3,865 f t beyond the threshold
and 97.5 ft to starboard of the edge of Runway 23, Auckland Inter-
national Airport, New Zealand. Geographic location: Lat. 37O 00’ 36” S ;
Long. 174O 47’ 29’‘ E. Elevation: 22 ft AMSL.
*

ACTION
3. O n the evening of 4 July 1966, investigation headquarters were
established at Auckland International Airport and the formal investiga-
tion was begun on the following morning. Thi! followed procedures
adopted by International Civil Aviation Organisation member States
experienced in conducting investigations into accidents involving large
modem transport aircraft and included formation of specialist groups,
under the direction and control of an Investigator in Charge, to obtain
factual information pertinent to structures, systems,. power plants, opera-
tions, flight data recorder, maintenance records, witness statements, and
human factors (aeromedical).

PERSONNEL
4.The investigation was led and directed by:
0. J. OBrien: Investigator in Charge; Chief Inspector of Accidents,
Department of .Civil Aviation, Wellington ;
E. F. Harvie: Inspector of Accidents, Department of Civil Aviation,
Wellington.
0
 5. Chairmen of the principal operations and technical p u p s were :
E. T. Kippenberger : Controller of Flight Operations, DCA, Welling-
ton;
W. R. Apperley : Regional Aircraft Surveyor, DCA, Auckland.

6. Other participating personnel were :

Accredited Representatives Acting as Specialist Consultants t o I.I.C.
M. J. Bates: Flight Safety Engineer, Douglas Aircraft Co. Inc., Long
Beach, California, U.S.A. ;
C. E. Richards : Supervisory Engineering Test Pilot, Federal Aviation
Agency (Western Region), Los Angeles, California, U.S.A.;
B. Cornthwaite : Technical Assistant to Engineering Manager, Air
New Zealand Ltd.;
I. H. Gemmell: DC-8 Training Captain, Air New Zealand Ltd.;
B. D. Dunn: Line Captain, New Zealand National Airways Corpora-
tion, ALPA/IFALPA.
Technical Advisers Appointed by Investigator in Charge
D. W. G. Keesing: Chief Pilot/Deputy Operations Manager, Air New
Zealand Ltd. ;
F. C. Hensley : Resident representative, Douglas Aircraft Field Service;
J. R. Herbein: Resident representative, Pratt and Whitney Field
Service;
J. W. Scott: Chief Inspector, Air New Zealand Ltd.;
J. R. Harding: Senior Servicing Inspector, Air New Zealand Ltd.;
D. B. Craig: Power Plant Foreman, Air New Zealand Ltd.

BRIEF CIRCUMSTANCES
7. With a complement of five persons aboard, the aircraft was
making the first take-off of a routine crew training flight. #Alloccupants
were seated on the flight deck. Shortly after rotation, the starboard wing
dropped, the aircraft failed to accelerate and gain height normally and
sideslipped inward until the wing tip struck the ground close to the
edge of the active runway. After impact (the aircraft slewed to Fort
and progressively disintegrated. Fire broke out in certain detached
portions but did not reach the flight deck which became separated
from the rest of the fuselage. Two occupants lost their lives, three were
seriously injured, and the aircraft was destroyed.

CASUALTIES
8. The following persons lost their lives:
Captain Dona1 McLachlan, DC-8 training captain, in command;
Flight Engineer Officer Gordon Kessell Tonkin, flight engineer.
The following persons were seriously injured :
Captain Bernard Joseph Wyatt, supernumerary flight crew ;
First Officer Brian Charles Ruffell, DC-8 co-pilot undergoing continua-
tion training;
First Officer Kenneth Allister Sawyer, supernumerary flight crew.
 FLIGHT CREW
9. Captain McLachlan, 46, joined Tasman Empire Airways Ltd. (now
Air New Zealand Ltd.) in 1947 with a background of 1,788 hours’ flying
experience accumulated principally with the RNZAF. I n 1951 he
obtained Air Line Transport Pilot Licence No. 67 and qualified for
line captaincy in 1958, thereafter commanding Short S.25, Short S.45,
Douglas DC-6, Lockheed 188C, and Douglas DC-8 aircraft. In mid-1965
he was among the first group of Air New Zealand captains to obtain
a DC-8 type rating and in September that year qualified as a conversion
instructor for that type. At the time of his death, Captain McLachlan
held valid ALTP and Flight Navigator Licences. I n a total of 17,966
hours’ experience, he had accumulated 16,178 hours with TEAL/Air
New Zealand Ltd. His experience on DC-8 aircraft totalled 497 hours
of which 69 hours had been obtained within the three months’ period
preceding the date of the accident. His last routine medical examination
had been undertaken in February 1966 when he was assessed fit without
waiver.
10. Flight Engineer Officer Tonkin, 33, joined Tasman Empire Airways
Ltd. in May 1949 as an engineering apprentice, subsequently obtaining
Flight Engineer Licence No. 39 in November 1958. During his term
of service with TEAL/Air New Zealand Ltd. he exercised his qualifica-
tions in Douglas DC-6, Lockkeed 188C, and Douglas DC-8 aircraft.
His DC-8 rating was obtained in September 1965 and at the time of
his death he had accumulated some 4,250 hours’ flight experience.
His licence was valid and he was medically fit.
11. First Ofiicer RufIell, 29, learned to fly in 1958 and obtained
Commercial Pilot Licence No. 935 before joining New Zealand National
Airways Corporation in 1960 as a trainee First Officer. Subsequently,
he obtained type ratings for Douglas DC-3 and Fokker F-27 aircraft
and qualified for ALTP Licence No. 316 in February 1965. In October
1965 he joined Air New Zealand Ltd. and a t the time of the accident
was undergoing continuation training as a co-pilot of DC-8 aircraft
in which he had accumulated 21 hours’ experience. First Officer Ruffell’s
total flying time is some 4,200 hours. At his last medical examination
in February 1966 he was assessed fit without waiver.
12. Captain Wyatt and First Officer Sawyer had no specific duties to
perform on the flight involving the accident.

AIRCRAFT
13. The aircraft involved in the accident was a Douglas DC-8 Series 52,
manufacturer’s Serial No. 45751, of New Zealand registry and markings
ZK-NZB, and owned and operated by Air New Zealand Ltd., a properly
constituted and certificated international carrier. The aircraft was built
in the U.S.A. by the Douglas Aircraft Co. Inc. in 1965 and was powered
by four Pratt and Whitney Model JT3D-3B turbofan engines each
capable of delivering 18,000 lb sea level static thrust.
14. Airframe time logged since new amounted to 2;275 hours. The
aircraft’s last “E” Check was completed on 1 July 1966 when airframe
hours totalled 2,248. A routine flight test was carried out immediately
afterward. Records detailing work completed and modifications embodied
during “E” Check and those pertaining to the flight test were examined.
c
 15. Engine running times were as follows:
Engine Total Hours Hours Since
Installation
No. 1 Serial No. 644650 1,850.50 26.55
No. 2 Serial No. 644617 1,829.00 480.00
No. 3 Serial No. 645185 925.05 925.05
No. 4 Serial No. 645103 1,322.20 1,322.20
16. A United Data Control (UDC) type FA542B flight data recorder,
with parameters covering compass heading, air speed, altitude and
vertical acceleration (G forces) measured against a time scale at one
second intervals on 5 in. inconel tape, was installed in the aircraft and
was in use at the time of the accident. The instrument was recovered
undamaged from the wreckage and a read-out obtained.
17. The last Maintenance Release issued in respect of the aircraft
was DSlOC No. 1846.

WEATHER
18. At the time of the accident, weather observed at Auckland Inter-
national Airport was as follows: Wind 240°T, 2 knots. Visibility 7 NM.
Cloud - cumulus, base 1,500 ft. Barometric pressure, 1,026 mb. Weather -
cloudy, light overcast with haze; adjacent showers. No turbulence was
reported in the area.

HISTORY OF THE FLIGHT

19. At 1537 hours on 4 July 1966, aircraft ZK-NZB advised Auckland
Tower that he would be starting engines in t y o minutes’ time in
preparation for a routine crew training flight within the Auckland
terminal zone and scheduled to last two hours. Purpose of the detail
was to afford First Officer Ruffell further co-pilot training on the DC-8
type aircraft. First Officer Ruffell occupied the left-front seat; Captain
McLachlan, the instructor, the right-front seat; Flight Engineer Officer
Tonkin the systems engineer’s station; Captain Wyatt the observer’s
seat behind the left-front seat; and First Officer Sawyer the navigator’s
seat.
20. The proposed start was delayed by late arrival of starting equip-
ment, but at 1550 hours ZB advised Tower that he was then starting
engines. At 1554 ZB was granted taxi clearance as far as the holding
position near the threshold of RW 23 and three minutes later was
cleared to line up on the active runway. Finally, at 1559, ZB was cleared
for take-off when ready. From a static start, the aircraft then began its
take-off roll from a point 230 ft forward of the threshold.
21. The aircraft made an apparently normal take-off roll without
observed deviation from the runway centreline and, at a point close to
the intersection of Speedway 4 and approximately abeam the passenger
terminal building, began rotation. Rotation appeared more rapidly
achieved and steeper than usual and the aircraft’s tail passed unusually
close to the runway surface. After becoming airborne, ZB appeared
briefly to maintain its original heading but not to gain much height.
Almost immediately, the starboard wing dropped and the aircraft began
turning to starboard while still in a nose-up attitude. Momentarily, the
nose appeared to be lowered and the aircraft to stop turning but the
starboard w F g continued to go down and the aircraft lost height by
sideslipping mward. The starboard wing 'tip then struck a grassed area
close to the edge of the runway and appeared to be worn down until
the aircraft pivoted about its nose at a fuselage-to-ground angle of
about 50'.
22. Fire broke out in the vicinity of the starboard wing root and the
aircraft rapidly began to disintegrate. Following nose impact, the entire
flight deck section broke loose and eventually came to rest inverted.
That portion of the wreckage carried the greatest distance by inertia was
found 1,070 ft beyond the point of wing tip contact. Fire fighting and
rescue services were immediately brought into action and the five
occupants extricated from the flight deck. Two crew members died,
three others were seriously injured, and the aircraft was destroyed.
23. No message or distress call from the aircraft was received by
Auckland Tower between start of the take-off roll and occurrence of
the accident. Twenty-five seconds elapsed between initiation of the
take-off roll and the time when the flight data recorder registered forces
indicative of severe disruption.

INVESTIGATION
Witness Accounts
24. Interviews were held with 12 eyewitnesses who, from various
positions, had seen more of events preceding the accident than others
whose testimony was considered much too limited in scope and value
to record. Eleven of those witnesses, employees of Air New Zealand
Ltd., were fully accustomed to watching DC-8 aircraft taking off on
both scheduled service and crew training flights and were technically
well qualified. The twelfth was an executive officer of the operations
division of an overseas airline and similarly well qualified.
25. None of the 12 witnesses saw the aircraft begin its take-off but
seven saw the aircraft at some stage of its roll down the runway. Three
of the seven heard a change in engine noise while one other person who
did not see the aircraft at all also heard it. All four associated it with
that to which they had become accustomed whenever simulated engine
failure on take-off had been resorted to on training flights.
26. Four of the seven witnesses noticed that the fan cascade doors of
No. 4 (starboard outer) engine were fully open when the aircraft passed
their viewing point. Three of these persons were watching from an
upstairs room, the fourth from a lower floor situation in the same
building. Two of the four also noticed that the primary cascade sleeve
of No. 4 engine was in the rearward, or reverse thrust, configuration.
Three of these persons expected, in consequence, to see the take-off
aborted. None of the seven witnesses noticed whether the fan cascade
doors or the primary reverser sleeve moved thereafter from the open
to the closed position.
27. The distance between the threshold of the active runway and the
point at which the fan cascade doors had been seen in an open configura-
tion was approximately 2,090 ft.
 2:). Six of the seven witnesses referred to watched the aircraft rotate
and five thought that this took place where they would have expected it.
All 4x considered that the aircraft's ground speed at rotation had been
not i i i n l . The point of rotation was some 440 f t along the runway beyond
thc \,icwing position of five witnesses and some 1,210 f t in advance of
thr position of the sixth. Five of the witnesses claimed that, during
rot:ition, the nose of the aircraft rose higher than usual, while all six
s t a t d that the tail cleared the runway by a very small margin.
20. The bulk of the evidence suggested that immediately after rotation
the aircraft momentarily maintained direction and remained laterally
levcl but that the starboard wing then began to move progressively
dov. rirvard without observable interruption while, simultaneously, the
airci.ift turned to starboard. I t then sideslipped inward until the starboard
wing tip struck the ground.
30. The majority view is that the aircraft never achieved a height
the ground greater than about 100 ft. No witness noticed whether
rhc .iilerons moved during the period between lift-off and wing tip
imp.wt. One person, however, stated that full left rudder was applied
thoiyli he had not actually seen the rudder moving into that configura-
tion. Full left rudder, he said, was kept applied at least until the wing
tip struck. At some undetermined moment after lift-off two persons
noticwl that the fan cascade doors of No. 4 engine remained open.
The\ did not see them shut but made the point that changing attitude
of tlie aircraft and increasing distance had made it difficult to keep
tIlc doors in view. Nearly all witnesses questioned said that the fuselage
had not assumed a horizontal attitude until after the wing tip had
struck the ground.
3 1. I n summary, witness accounts suggested that-
(1) The take-off roll was normal; 4

( 2 ) -4change of engine noise, symptomatic of a reduction of power

on one engine, occurred during the take-off roll ;
( 3 ) T h e fan cascade doors and primary reverser sleeve of No. 4 engine
were in a reverse thrust configuration during part of the
take-off roll;
(4)-As the aircraft rotated, the nose rose higher and the tail passed
closer to the ground than usual;
(5' .Ilmost immediately after rotation, the starboard wing dropped and
continued to go down while, simultaneously, the aircraft began
to turn to the right;
(6' The aircraft sideslipped inward until the wing tip struck the
ground;
(7' T h e fan cascade doors of No. 4 engine were open for an
undetermined time after the aircraft left the ground;
( 8 -After lift-off, the rudder was deflected toward the left, but there
was no apparenit aileron movement.
rVreL-1.7ge Distribution
32. -At a point 3,865 f t from the threshold and 97.5 ft to the right of
the edge of RW 23, first indications of starboard wing tip contact with
the gTmnd were observed.
c

33. Lengthy ground scars between wing tip and nose radome impact
points, positions along that line at which certain components had
become detached, and those where structural sections or important
components, including, engines, finally came to rest were plotted to
provide a diagrammatic presentation of the wreckage distribution from
which analysis could be made and specific information obtained.
34. The distance covered by the aircraft from rotation point to point
of wing tip contact was approximately 1,345 ft. The latter point
represents an angular deviation of 64O from the point of rotation on
the runway centreline. Thereafter, the track of the aircraft continued
toward the right for 480 f t until the nose radome struck at a point
where angular deviation reached 12O. At the point where the wreckage
trail ended, angular deviation was 21O.
35. Well defined ground scars extended almost continuously as far
as nose radome impact point. A determination of the angle of bank
assumed by the aircraft after being in contact with the ground for 70 f t
was obtained from a deep scar displaying a sharply cut inner edge.
Mter allowing for mainplane dihedral, it was determined that in that
area the aircralft had been banked between 43.3O and 46O to Starboard.
36. Prior to nose radome impact certain components became detached
and were found in the order listed: starboard navigation light cover-
glass; starboard outer wing tank inspection panel; starboard cascade vane
from No. 4 engine; starboard wing tip fairing; outboard pylon strut,
No. 4 engine; and a section of starboard wing tip measuring approxi-
mately 5 in. chordwise and 30 in. spanwise.
37. In this report a detailed description of the wreckage and its
distribution is considered unnecessary in that much of it is of no relevance
to the cause of the. accident. Reference, however, will later be made to
specific damage of importance in the technical examination. I t may be
briefly noted, nevertheless, that the entire flight deck zone became
detached from the neighbouring structure and was untouched by fire;
that Nos. 1, 3, and 4 engines became dislodged from their pylons during
disintegration; that the passenger cabin aft of Station 620 was gutzed
by fire; and that the landing gear selector lever was‘ found “down”
and one set of main gear wheels, severely damaged, was found down
and locked.
38. The wreckage, including instrumentation, controls and systems as
found, was fully documented and an appropriate photographic record
obtained.
39. In summary, evidence obtained from the wreckage trail showed
that the aircraft first struck the grouhd while banked a t about 40° to
starboard and that shortly afterward it pivoted to port about its nose,
subsided rearward and disintegrated, fire rapidly engulfing the passenger
accommodation area aft of the mainplane leading edge junction points.
The physical state of the wreckage confirmed descriptions of the air-
craft’s behaviour provided by witnesses.
Flight Data Recorder Read-out
40. The aircraft’s UDC FA542B flight data recorder, provided with
parameters registering compass heading, air speed, altitude and vertical
acceleration (G forces) on 5 in. inconel tape a t one-second time intervals,
was recovered intact, calibrated against laboratory standards and
interpreted.
-
 ,

41. I n the table below, information provided by altitude and vertical

acceleration channels is omitted. With altitudes covering an extremely
narrow range very close to the runway, ground effect, local pressure
differentials and other factors make accurate interpretation so difficult
that figures obtained cannot be regarded as wholly reliable. Information
which might have been obtained from either parameter was, however,
of no material importance in establishing the cause of the accident.
42. Against the time scale in the table below, information relative to
the aircraft’s status, second by second, has been included for purposes of
explanation. Take-off weight was 206,000 Ib and in that configuration
V, (see para. 87 for definition) had been assessed by the flight crew
prior to take-off as 103 knots and V, (see para. 89 for definition) as
118 knots. A reference card carried by the active aircrew displayed this
information which was checked after the accident and found to have
been correctly computed.
43. Flight data recorder information covering compass heading and
air speed channels is tabulated below :
Time from Compass Air
Start of Heading Speed
Take-off in Degrees in Aircraft Status
in Seconds Mag. Knots
1 233.5 Not Aircraft rolling and accelerat-
2 233.75 reliable ing.
3 234-5 until
4 235.0 aircraft
5 235.0 has
6 235.0 reached
7 235.0 an
8 235.0 air
9 235.0 speed
10 235.0 of
11 235.0 approx.
12 235-0 100 knots
13 235.0
14 235.0
14.2 235.0 103 VI call speed.
15 235.0 107
16 235.0 112
17 235.0 118 V, call speed; rotation begun.
18 2345 120 Half degree deviation to port.
19 234.5 123 Should be airborne.
20 23+5 124.5 Maximum air speed achieved.
21 234.5 124 Lack of acceleration.
22 235-0 121
23 236.0 119 Heading starboard, speed de-
creasing.
23.5 237.0 118 Initial wing tip impact.
24 239.0 116 Further decrease in air speed,
aircraft heading further star-
board.
25 2M5 114 Nose impact; heavy disturb-
ance registered on all chan-
nels.
26 255.0 110 Aircraft disintegrating.
 ,

44. Read-out indicates that shortly after rotation the aircraft headed
to starboard and failed to accelerate normally; it confirms behaviour
of the aircraft as described by witnesses.
45. Significant information obtained from the read-out is discussed
under the heading “Analysis”.
Survivors’ Accounts
46.Following medical advice that they were considered well enough
to be interviewed, Captain Wyatt and First Officer Ruffell were
questioned independently on 6 July 1966. Neither survivor had seen
or spoken to the other since the occurrence of the accident.
47. Captain Wyatt said that-
(a) The initial take-off roll was normal;
(b) Before V, was reached he stood up;
(c) He heard VI called;
(d) He saw Captain McLachlan operate No. 4 power lever to simulate
failure of No. 4 (starboard outer) engine; Captain McLachIan used the
spoiler disarm extension (a short bar projecting laterally and toward the
right from No. 4 power lever) to “snap” the power lever back very
rapidly;
(e) A slight yaw resulted but First Officer Ruffell corrected it;
(f) He resumed his seat;
(g) He could not remember whether he heard V, called but thought
that, initially, rotation was rather rapid; the rate of rotation, he believed,
was then checked and he had assessed pitch attitude to be about loo
ANU;
(h) During rotation he had felt no severe yaw but became alarmed a
short time afterward when, with the aircraft banked some 3 5 O to star-
board, he saw the ground through the co-piIot’s side window;
(i) He stood up, felt the aircraft sinking and noticed bank increasing;
believing that the aircraft would strike the ground, he threw himself
to the floor behind the left-front seat;
(j) He had not noticed whether any cockpit indicator light had
become illuminated;
(k) He felt the starboard wing tip strike the ground;
(1) No conversation took place between the two pilots;
(m) He had not noticed whether there had been any indication of
failure of any engine other than of No. 4 whose failure had been
simulated ;
(n) He had not noticed whether No. 4 thrust brake lever had at any
time been in a reverse thrust position.
48. First Officer Ruff ell said that-
(a) He had expected to be required without warning to cope with
simulated failure of one engine at VI;
(b) The take-off roll had been normal;
(c) He remembered VI being called, thereafter feeling the aircraft
tending to yaw but correcting this with application of rudder;

11
 (d) He assumed that Captain McLachlan had pulled the power off
No. 4 engine but did not look at the pedestal to determine whether,
in fact, that had been the case;
(e) He had had to apply forward pressure on the control column to
prevent nose wheel scrubbing but had kept the aircraft straight;
( f ) He could not remember whether he had heard V, called or
whether he had glanced at his own air speed indicator;
(g) On rotating the aircraft he had “felt something happen”, as
though the instructor might have simulated failure of No. 3 en,’vine as
well as of No. 4;
( h ) After rotation he had had to apply full left rudder to counteract
yaw and though initial climb had been normal he believed that air
speed had been low;
(i) He had been unable to hold the aircraft laterally level and could
not recall whether he had moved the control column forward and/or
had applied aileron to bring the wing up;
(j) He could not remember whether Captain McLachlan had taken
any specific action after rotation but “thought he was doing something
with the throttles” (power levers) ;
(k) Neither pilot spoke to the other following rotation;
(1) H e had not felt the stick shaker operate nor heard any stall
warning.
Technical Investigation
49. Principal objectives of the technical investigation (which included
obtaining an appropriate photographic record, complete documentation,
calibration of specific instruments and rig tests of systems components)
were-
(i) To determine whether any failure, malfunction, or defect in the
airframe, flight or systems controls, instrumentation or
systems components had caused or contributed toward the
occurrence of the accident;
(ii) To ascertain whether failure, actual or simulated, or loss of
thrust occurred in any engine other than in No. 4;
(iii) To determine whether any unexpected or unusual occurrence
had accompanied simulated failure of No. 4 engine;
(iv) To determine, if possible, power output, fuel flow, and reliability
of services in respect of all four engines;
(v) To determine whether any mishandling of flight or systems
controls might have contributed toward or caused a hazard-
ous in-flight situation;
(vi) To determine whether any other factor might have caused the
accident.
50. Information obtained during the technical investigation is docu-
mented in detail elsewhere, but unless of specific relevance is excluded
from this report. A brief summary of important determinations, however,
now follows:
51. Structural Integrity-No evidence to show that the accident
resulted from any structural failure or defect likely to produce a critical
in-flight situation was found. On the flight preceding that on which
* I
the accident occurred, the aircraft had behaved normally in all respects.
There was no evidence to show that the aircraft had not been airworthy
at the time when it began its last take-off.
52. Cockpit Instrumentation and Control Status-Information of
importance in the technical investigation included the following cockpit
indications :
Fuel quantity totaliser: 70,000 lh remaining in tanks.
Fuel flow indications :
No. 1 engine-231 lb.
No. 2 engine-233 lb.
No. 3 engine-291 lb.
No. 4 engine-219 lb.
Fuel selectors: All ON.
EPR (engine pressure ratio) indicators:
NO. 1 - 1.24.
No. 2-1.11.
No. 3 - 1.07.
NO. 4-0.96.
L.H. hydraulic pump switch : In “bypass” configuration.
Hydraulic pressure indicator: 3,100 p.s.i.
Flap selector: Handle out of detent and broken; indicating loo flap
extension.
Gust lock: OFF.
Doppler reading: Ground speed 123.5 knots (Xote: Doppler reverts
to memory mode when bank angle exceeds 20O.)
53. Flight Controls-(a) Ailerons: Aileron power was selected ON;
the RH hydraulic pump was switched OX and was operational; hydrau-
lic pressure was available to the entire sy5tem; aileron installation was
normal in all respects, with inboard and outboard units free to move
throughout their normal travel ; power cylinders and control valves
were operable; cockpit control wheels were free to, move; there was
no evidence of pre-impact defect in the aileron system as a whole.
(b) Spoilers: L H and RH spoilers were found down and locked; if
any, less than 45O aileron control wheel movement was being applied at
the time of impact; the flight spoiler system was capable of normal
operation.
(c) Flaps: There were no indications of flap failure or malfunction;
prior to impact flaps were extended not less than loo.
(d) Rudder: The power lever was in its appropriate detent, and
indicating ON.
(e) Horizontal Stabiliser: Positioning mechanism suffered severe im-
pact damage and no thoroughly reliable information was obtained.
54. No evidence of pre-impact failure, defect, or malfunction in the
flight controls or allied systems which might have accounted for the
observed behaviour of the aircraft after lift-off was found.
55. Instrzimentation-Both horizon flight directors (including pitch
trimmers) and air speed indicators were calibrated, functionally tested,
and found to operate within specification tolerances.
 56. Of EPR readings, the Douglas DC-8 Systems Operation Manual
notes that “the indicator pointer and digital counter tend to remain in
their last indicating position if the system electrical power source is
shut OW’.

57. The doppler reading of 123.5 knots was considered to be a reliable

indication of the ground speed prevailing immediately before the aircraft
exceeded an angle of bank of 20° to starboard.

58. Fuel Quantity and Quality-Ample fuel for the intended flight
was carried. All Avtur fuel carried comprised part of Shell batch A/H 149.
Immediately after the accident the bulk supply, the tanker and trailer
involved in the most recent refuelling detail for ZK-NZB were quaran-
tined and samples uplifted for laboratory analysis. All samples tested
conformed fully with Avtur Specification DERD 2494.

59. Power Plants-(a) N o . 1 (port outer) : This engine was thrown

off its pylon by inertia forces and was not dislodged in shear by ground
contact. Gouges and tip bending were present in certain blades of the
first fan stage. The turbine appeared undamaged. A fan-shaped area of
burnt grass immediately aft of the tail pipe indicated that the turbine
had been rotating and discharging hot gases when and after the engine
came to rest. EPR reading, 1.24; fuel used, 231 lb.
(b) NO.2 (port inner) : This engine remained on its pylon until, when
almost at rest, it became dislodged beneath the wing. Heavy impact
forces damaged inlet guide vanes, main compressor stages, and portions
of the turbine section. Damage to fourth stage turbine buckets was
incurred by arrestation at high rotational speed. A large fire in the hot
section was responsible for the conflagration which ultimately gutted
the fuselage aft of the leading edge junction point. EPR reading, 1.11;
fuel consumed, 233 lb.
(c) N o . 3 (starboard inner) : This engine became detached from its
pylon. Insulation material from the cabin walls was ingested but this
did not occur at the point where the engine came to rest. It was
established that the engine was still rotating at high speed while the
cabin was disintegrating and that insulation material was ingested before
the engine left its pylon. The turbine was undamaged. A small fire
occurred in the accessory section. The fuel control and main engine
pump were recovered intact and during their removal for test fuel was
found present a t all connection points. All filters were clean. Power
plant control cables in the pylon area showed impact-type tensile breaks.
Controls from those breaks to the fuel control were correctly rigged.
Nothing was found to indicate any interruption in the normal flow of
fuel to No. 3 engine. There was clear evidence of high speed rotation at
the moment of impact. EPR reading, 1.07; fuel used, 291 Ib.
( d ) No. 4 (starboard outer) : Shortly after rhe aircraft began to slew
to port, this engine made heavy impact and broke away from ilts pylon.
Severe damage was incurred by the forward section, many compressor
blades being broken or bent in a direction contrary to that of rotation.
It was not possible to determine turbine speed at the time of impact.
A single fan cascade vane was found close to the point where the engine
struck the ground and this might have been regarded as an indication
that the cascade doors were open at the time of its detachment. Careful
examination, however, revealed that at the time of engine impact the
doors were, in fact, stowed and locked; that is, the engine had not been
armed for thrust reversal when the vane became detached. ‘The fuel
control was recovered intact and a functional check made of the reverser
pilot valve which operated normally. EPR reading, 0.96; fuel used, 219 Ib.

60. It was not possible to establish precisely how much thrust was
being developed by any of the engines a t the time the aircraft struck
the ground. Physical evidence, however, pointed toward normal power
plant operation and an entire absence of mechanical trouble. There was
no evidence of failure of No. 3 engine which, in view of a statement by
First Officer Ruffell (see para. 48 (g)), might have been held suspect.
Examination had been required to determine whether or not simulated
failure of No. 4 engine had been accompanied by actual failure of No. 3,
even if the latter’s fuel consumption virtually ruled out that possibility.

61.L H Hydraulic Pump-A post-accident check showed that the

cockpit switch was in the “bypass” configuration. Later investigation,
however, showed that the switch had assumed that position either
through severe disruptive forces following nose impact or at a time
when rescue operations within the flight deck area were being carried
out and certain damage was incurred. I n any event, the position in
which the switch was found could not be reconciled with other and
more reliable evidence and was ultimately considered to have no
relevance to the cause of the accident.

Fan Cascade Doors

62. Paragraph 26 mentions that three persons in an upstairs room and
one in a lower-floor room facing the runway saw the fan cascade doors
of No. 4 engine in an open configuration during the take-off roll.
Paragraph 30 notes that two persons saw them in the same condition
at some undetermined moment after the aircraft was airborne. It was
necessary to determine how reliable that evidence was and a demonstra-
tion with another DC-8 aircraft was arranged accordingly. The aircraft
was first stopped at a point on the runway where witnesses had seen
the open doors. The doors were opened and shut several times to allow
observers to form their own opinions. The aircraft then returned to the
threshold and began a take-off roll and when it had reached the
aforementioned point the cascade ‘doors of No. 4 engine were opened
and left open until, ultimately, the take-off was aborted. These trials were
made a t a time of day about one hour earlier than that on which the
accident had occurred but in virtually identical conditions of natural
light. They proved quite conclusively that it was very easy to distinguish
an open-door configuration against a closed one, there being a marked
difference in the colour and appearance of the relevant portion of the
engine nacelle whenever the fan cascade doors were open. It was there-
fore concluded that what was reported.to have been seen was, in fact,
seen: the fan cascade doors of No. 4 engine had indeed been open during
part of the take-off roll and they had remained open for some time after
the aircraft had left the runway.
15
Engine Power Levers and Reverse Thrust Controls
63.The reverse thrust control (properly known as the thrust brake
lever) for each engine of the DC-8-52 consists of a short lever, some
4 in. long, attached through a pivotal point to the top of each power
lever (sometimes called a throttle). I n a take-off thrust configuration,
the power lever is in a forward position, its associated thrust brake lever
pointing forward and downward beyond its knob. To reduce take-off
thrust or to achieve a state of forward idle, the power lever is moved
rearward, its associated thrust brake lever still remaining pointed forward
and downward beyond the knob. Note that with forward or rearward
movement of the power lever, the thrust brake lever is not moved.
64. Whenever it is desired to apply reverse thrust - say, to achieve a
shorter landing roll - two distinctly separate movements must initially
be made: (i) the power lever must be moved back until it reaches the
pedestal backstop; (ii) the thrust brake lever must then be Zifted upward
into the reverse idle detent. Note that the thrust brake lever cannot be
raised unless and until the power lever is moved fully rearward.
65. Lifting the thrust brake lever into the detent actuates a pilot valve
which then arms the engine for application of reverse thrust when
needed. I t causes the fan cascade doors to open fully; the sleeve covering
the primary cascades in the turbine section to slide rearward and expose
the primary cascades; and the reverser buckets to close and seal off the
tail pipe from which the jet efflux normally emerges.
66. Application of reverse thrust is then achieved by moving the thrust
brake lever further backward. Fan stage air is thereupon discharged
forward through the open fan cascade doors; jet efflux gases strike the
reverser buckets in the tail pipe and are discharged forward through the
open primary cascade doors exposed by the rearwardly positioned sleeve.

67. Though No. 3 engine nacelle could not clearly be seen by witnesses,
it was considered that No. 3 engine had not been armed concurrently
with No. 4 for thrust reversal. If it had, so would Nos. 1 and 2, and
intention to abort the take-off would have been apparent. No change
of engine noise indicative of an immediate abort had been heard and
the aircraft had continued its take-off roll and been duly rotated. Since it
is apparent that an aborted take-off had not been intended, it had been
unexpected and unusual to see No. 4 engine armed for reverse thrust.
I t therefore became necessary to determine whether, in simulating failure
of No. 4 engine - a common enough exercise on crew training flights -
the arming of No. 4 into a reverse thrust configuration had been incurred
by mischance.
68. One of the survivors, Captain Wyatt, had taken particular notice
of the rapidity with which Captain McLachlan had moved No. 4 power
lever backward. He had also heard the sharp click made by the lever
when it had been brought against the pedestal backstop. Furthermore,
he had noticed that Captain McLachlan had used the spoiler disarm
extension - a small bar protruding horizontally toward the right from
the approximate central point of the power lever - to “snap” the lever
back.
I C
 69, Experiments were conducted with the power levers of another
DC-8 aircraft, ZK-NZC. With the friction lock released and with No. 4
power lever in the take-off thrust position, the lever was held by the
spoiler disarm extension and moved very rapidly rearward until it struck
the backstop. Inertia force thereby incurred caused the thrust brake
lever to move from its normal forward idle position into one of reuerse
idle. Simultaneously, No. 4 reverse indicator lights came on to show that
the fan cascade doors had opened, the primary cascade sleeve had moved
rearward, and the reverser buckets had closed off the tail pipe. A physical
inspection of those components confirmed that reversal light indications
were reliable. The engine was armed for reverse thrust.

70.Another test was made, this time with the knob of No. 4 power
lever held between thumb and forefinger in such a way as to leave the
thrust brake lever, pivoted about the knob, free to move. Though it was
found more difficult to “snap” the power lever rearward as rapidly as
before, sufIicient inertia was produced to cause the thrust brake lever
to move into the reverse idle detent.

71. Both tests were repeated several times by various personnel and
clearly demonstrated that if No. 4 power lever were moved rearward
sufficientIy quickly the thrust brake lever would move into a reverse idle
position and the engine become armed for thrust reversal.

72.Next, the power lever of No. 3 engine was held in the manner
described in para. 70 and moved rapidly backward. Identical results
were obtained.
73. Applied at the pivot point, the force required to permit the thrust
brake lever of each engine to enter its reverse idle detent was then
measured. The following results were obtained :
No. 1 thrust brake lever: 5.1 Ib.
No. 2 thrust brake lever: 5.75 lb.
No. 3 thrust brake lever: 5.9 Ib. ,
No. 4 thrust brake lever: 4.75 Ib.
74. Following these experiments, the power lever cable systems for
the engines of ZK-NZC were checked by meter for rigging tension and
found to be within specification tolerance.
75. Trials were also conducted with another DC-8 aircraft, ZK-NZA.
Tension meter readings were as follows:
No. 1 thrust brake lever: 5.0 lb.
No. 2 thrust brake lever: 3.75 Ib.
No. 3 thrust brake lever: 4.25 Ib.
No. 4 thrust brake lever: 4.25 Ib.
These figures were likewise within rigging tolerance.
76. On account of impact damage incurred, it was not possible to
conduct tests with the power levers of the aircraft involved in the
accident, ZK-NZB.

17
Previous Incident
77. Since simulated engine failure necessitates physical rearward move-
ment of the power lever, and since experiments had shown that if that
movement were made very rapidly the thrust brake lever could some-
times, but not always, move into the reverse idle configuration, it was
considered essential to determine whether any case of unwanted reverse
thrust had previously occurred. Inquiry was immediately made of all
Air New Zealand DC-8 flight crews and arrangements made to obtain
information from DC-8 pilots in other parts of the world.

78. Inquiry brought forward two Air New Zealand first officers who,
while on the flight deck together and with Captain McLachlan as
instructor, had observed an incident involving, as far as one of them
could remember, aircraft ZK-NZB on 2 April 1966. This had occurred
during a training flight.

79. One of the officers, an observer at the time, reported that during
a “touch and go” landing Captain McLachlan had simulated faiiure
of No. 4 engine at approximately V,. When moving the power lever
rearward with his thumb and index finger, the instructor had “inadver-
tently flicked” the thrust brake lever toward the reverse idle detent
whereupon the reverser doors and buckets had become activated. The
first officer flying the aircraft had been unable to maintain directional
control with full left rudder and the aircraft had yawed toward the right.
I t was quickly seen that “the engine door and thrust brake lights were
illuminated for No. 4 engine and the lever was promptly returned to
the forward idle detent by Captain blclachlan. A normal three-engine
climb-out then followed”.

80. The first officer flying the aircraft stated that following a simulated
failure of No. 4 engine after VI the aircraft had, on beqoming airborne,
yawed violently to starboard. He had applied full left rudder and
sufficient aileron to level the wings but could not prevent yaw. Speed
had been 130-140 knots and to improve directional control he had
lowered the nose and maintained a lateral attitude not exceeding loo
angle of bank. When it was discovered that the failed engine was in a
state of reverse thrust, the thrust brake lever was moved forward and
a three-engine climb-out was made.

81. Paragraph 79 records that the witness stated that Captain McLach-
Ian had “inadvertently flicked” the thrust brake lever toward the reverse
idle detent. Post-accident experiments, nevertheless, showed conclusively
that the Ithrust brake lever could sometimes be caused by inertia forces
alone to move toward or into the reverse idle detent if the power lever
were moved very rapidly reanvard.

82. Inquiries made of Douglas DC-8 pilots throughout the world have
so far brought to light no case of inadvertent incurrence of reverse thrust
through the circumstances described. Neither, similarly, have the manu-
facturer’s test pilots nor FAA certification pilots or engineers experienced
it.
Precautionary Measures Instituted
83. Whenever, during the course of an investigation, any circumstance
or physical condition capable of producing an unexpected, adverse or
hazardous situation either on the ground or in flight is discovered - and
this applies whether or not what is discovered appears to have caused
or contributed toward a particular accident being investigated - aviation
authorities, manufacturer, and operators of the type everywhere are
immediately alerted so that they may institute measures to guard against
such circumstances or condition themselves. This sort of advice is
regarded as a precautionary safety measure and demonstrates only one
of the many recognised values of thorough investigative practice.
84. Three hours after experiments had yielded the results described,
arrangements had been completed for the immediate notification of
aviation authorities abroad, the Douglas Aircraft Co., all operators of
DC-8 aircraft, and the Department of Civil' Aviation in New Zealand.
At the time when this alert was promulgated, it was by no means certain
whether the condition shown to be possible of incurrence had, in fact,
constituted the basic cause of the accident. The message was primarily
intended to put all concerned on their guard.
85. As soon as the Douglas Aircraft Co. had received notification,
it conducted experiments of its own and promptly reported back that
it had achieved results similar to those found by trial in the New
Zealand investigation.
86. Immediately it appeared probable that unintended incurrence of
reverse thrust had constituted the basic cause of this accident, the
Minister of Civil Aviation was informed so that he might make a
public statement, principally to assure the travelling public that what
had occurred could not happen on scheduled passenger carrying flights,
to dispel rumour and to refute irresponsible speculation. Subsequent
intensive investigation confirmed this interim finding.

Definitions of Speeds
87. DecisionlRefusal Speed ( V I ) , in terms of calibrated air speed,
is the speed at which the pilot, having recognised sudden failure of the
critical engine, can, by use of primary aerodynamic controls alone,
adequately control the aircraft and proceed safely with the take-off
using average piloting skill or, alternatively, abandon the take-off and
bring the aircraft to a stop within the acceleration-stop distance available.
88. Minimum Control Speed (V,,,) , in terms of calibrated air speed,
is the speed at which, when sudden and complete failure of the critical
engine occurs at that speed, it is possible to recover control of the aircraft
with the engine still inoperative and to maintain it in straight flight at
that speed, either with zero yaw or with an angle of bank not in excess
of five degrees.
89. Rotation Sfieed (V,) is the calibrated air speed at which, during
the take-off roll, the aircraft is rotated about its transverse axis prior to
lift-off. It must not be less than V1 nor less than V,,,.
Applicability of Speeds to Z K - N Z B
90. Based on a gross take-off weight of 206,000 Ib, air temperature of
1Ioc, zero wind, 25O flap extension and D-1 rating (constant EPR
setting for all engine bleed conditions) a t sea level, the following speeds
had been calculated - and subsequently established to have been correct -
by the flight crew as being applicable to ZK-NZB a t the time of its
last take-off :
VI: 103 knots.
Vmca:113 knots.
V,: 118 knots.
91. I n response to a request for information, the Douglas Aircraft Co.
provided the following values for V,,, in respect of ZK-NZB under the
conditions outlined in para. 90 but with No. 4 engine in the reverse idle
configuration :
Wings level, V,,, with power control: 141 knots.
Wings level, V,,, with manual control: 147 knots.
92. V,,, for ZK-NZB in a banked configuration, a status achieved
very soon after lift-off, and with No. 4 in reverse idle, was:
Banked 5 O to starboard, V,,, with power control: 156 knots.
Banked 5 O to starboard, V,,, with manual control: 162 knots.
For every degree of unfavourable bank, an increase of approximately
3 knots over the wings-level V,,, is incurred.

ANALYSIS
Airframe, Systems, Instrumentation Integrity
93. There was no evidence to show that the entire airframe, the fliqht
controls, various systems and instrumentation were not airworthy and
fully operational before the aircraft first made impact with the ground.

Power Plant Integrity

94. Simulated failure of No. 4 engine was initiated by Captain
Mclachlan. Captain Wyatt had seen him move No. 4 power lever
rapidly rearward as far as the backstop; First Officer Ruffell had
detected a tendency of the aircraft to yaw; witnesses had heard a sound
typical of engine spooldown on simulated failure; and witnesses had
seen the aircraft with the fan cascade doors of No. 4 engine open - a
mechanical condition not possible unless No. 4 power lever had first
been placed in a flight idle position, i.e., moved fully rearward.
95. Failure of any other engine, either simulated or actual, is dis-
counted. Captain Wyatt had not seen Captain McLachlan simulate
failure of any engine other than No. 4; First Officer Ruffell had not
felt or suspected indications of a second failure while the aircraft was
still on the runway; physical evidence disclosed the presence at impact
of considerable power on Nos. 1, 2, and 3; no sound typical of engine
spooldown, other than that associated with No. 4, had been heard by
witnesses.
 96. Close examination of No. 4 engine disclosed no evidence to show
that prior to ground impact it had been in any way mechanically
defective.
97. Prior to impact, Nos. 1 and 2 engines had used almost identical
amounts of fuel, viz, 231 and 233 Ib respectively. It was not unusual
that No. 3 had used 291 Ib, since it had been started before any other
and then accelerated to provide bleed air €or starting Nos. 1, 2, and 4.
With a consumption of 219 Ib, No. 4 had used least fuel of all and
this was consistent with consumption expected from an engine in which
simulated failure had been incurred during the take-off. Quantities
consumed by Nos. 1, 2, and 3 are consistent with no failure or power
loss having occurred before the aircraft struck the ground.
98. I t is determined, therefore, that prior to impact all four engines
were operating strictly in accordance with the control commands that
had been transmitted to them.

T i m e of Simulated Failure
99. Simulated failure of No. 4 was initiated at or immediately after
the calling of VI. First Officer Ruffell would have had his right hand
on the four power levers until V1 was called and in consequence would
have felt a movement of No. 4 power lever had Captain McLachlan
attempted to pull the power OR earlier.

Indications of Reverse Thrust

100. No. 4 was armed for reverse thrust - i.e., the fan cascade doors
had opened, the primary cascade sleeve had moved rearward and the
reverser buckets had closed off the tail pipe -within two seconds of initia-
tion of simulated engine failure. Independently situated witnesses had seen
that the fan cascade doors were open after V1 point had been reached;
two witnesses had seen the same doors in an open stqte at some moment
after lift-off; behaviour of the aircraft after it had become airborne was
consistent with what would be expected with No. 4 engine in a condition
of reverse thrust.

Unintentional Arming for Reverse Thrust

101. Tests made with two other Air New Zealand DC-8 aircraft and
trials conducted by the Douglas Aircraft Co. in the U.S.A. showed that
when in some, but by no means all, cases a power lever had been moved
very rapidly rearward, its associated thrust brake lever could, through
inertia forces alone, move toward or into the reverse idle detent. In some
cases, this condition had been incurred with comparative ease, in others
with difficulty, but in all instances rigging tensions of relevant control
cables had been within specification tolerance. In view of the rapidity
with which Captain McLachlan had moved No. 4 power lever back,
and of the fact that he had used the spoiler disarm extension to make
that movement, there can be little doubt that inertia forces caused
the thrust brake lever to enter the reverse idle detent. The matter of
rapid power lever operation is referred to in paras. 123-127.
 .
102. Technical examination, functional checks, and dismantling of the
fuel control and its associated reverser pilot valve, together with inspec-
tion of No. 4 engine controls, disclosed no evidence that arming for
thrust reversal had been caused by any mechanical failure or defect.
Had either existed, it might have been impossible for Captain McLachlan
to bring the engine out of reverse thrust. The evidence showed that he
had been able to do so.

Rotation
103. Like certain other types, the modern jet transport aircraft must
be flown very much 'Laccordingto the book", certain actions having to
be pexformed at specifically defined air speeds from which no significant
deviation can be permitted. "Near enough is good enough" simply will
not do. Hence, in every individual case (for gross weight, temperature,
altitude, etc., always vary) both V1 and V,, for example, are computed
prior to take-off and are called as soon as they are reached. Additionally,
it is a routine requirement for the pilot to monitor his own air speed
indicator and to be ready to act if no call happens to be given.
104. Flight recorder traces showed that rotation was initiated at the
pre-computed speed, 118 knots, and that slight yaw and simultaneous
reduction in the rate of acceleration occurred immediately afterward. The
fact that neither Captain Wyatt nor First Officer Ruffell could remember
whether they had heard V, called is considered unimportant. Had
V, not been called, it would have been contrary to normal practice but
supervisory pilots occasionally deliberately omit it to determine whether
the trainee is complying with the requirements for monitoring his ASI.
Whether or not V, was called was of no relevance to the cause of the
accident.
105. Evidence disclosed that rotation was more rapid than usual,
calculations suggesting that it was completed in' about two seconds as
against a more usual three. This could have accounted, in part, for the
observed close passage of the tail to the ground after rotation was
started. However, a reduction in the rate of acceleration after lift-off
(a consequence of No. 4 engine being in reverse thrust) is a more
probable reason and would account also for rapid angular rotation and
initially steeper-than-usual nose-up pitch.
106. The amount of reverse thrust developed by No. 4 engine at and
just after rotation is also of relevance. The Pratt and Whitney JT3D-3B
engine has been timed to take 14 seconds to spool down from take-off
thrust to forward idle. When or immediately after VI was reached,
No. 4 engine was cut and rotation was begun barely 2.8 seconds later.
With power pulled off No. 4 and the simultaneous unintended movement
of the thrust brake lever, the engine would take about two seconds to
arm itself for thrust reversal. It may be claimed conservatively that
2.8 seconds after power was pulled off No. 4, between 50 and 60 percent
of take-off thrust would remain in the downspooling engine. This is
an appreciable amount and would have been effective at and just
after rotation. Reverse thrust is capable of producing nose-up pitch and
that incurred in this case may well have contributed toward the steeper-
than-usual pitch observed.
99
 107. It is considered that the oiserved rapid angular rotation and
initially steep nose-up pitch - unus;iI, but not unknown in training -
did not contribute toward the cautr of the accident.
Y a w , Roll, and Sideslip
108. When No. 4 power lever v a moved back to simulate engine
failure, the aircraft tended to yaw o: starboard but this was effectively
countered by First Officer Ruffell who said that he had applied left
rudder and moved the control column appropriately further forward.
When, however, the aircraft was rotated, a very different situation
developed. The unusual amount of yaw made the pilot wonder whether
Captain McLachlan had simulated failure of No. 3 engine as well as
of No. 4 and it became neczssary for him to apply full left rudder to
try to prevent it. This appreciable yaw was undoubtedly caused by the
amount of reverse thrust being delivered by No. 4 engine in the initial
stages of spooldown.
109. Thrust imbalance from a condition of.asymmetric power produces
a yawing moment dependent upon the thrust imbalance and upon the
lever a n n of the force. Deflection of rudder creates a side force on the
tail and contributes a yawing moment to offset that due to thrust
imbalance. With Nos. 1, 2, and 3 engines delivering take-off thrust and
No. 4 in forward idle, the rudder is powerful enough to counteract the
yawing moment present. At low air speeds, however, and with Nos. 1, 2,
and 3 delivering take-off thrust hut with No. 4 in some condition of
reverse thrust, application of rudder may not be effective in wholly
overcoming the yawing moment incurred in that configuration.
llO.The flight recorder trace shows that after V, the aircraft’s
heading initially changed half a degree to port, the new direction being
maintained for a full three seconds, after which divergence to starboard
began and continued. It should be noted that angular divergence from
rotation point to point of wing tip impact was only Sit0, but that the
wing tip struck while the aircraft was banked at some 40° to
starboard. It therefore appears that since directional control was so
comparatively well maintained, the rudder was dose to being fully
effective in counteracting the prevailing thrust imbalance. It is obvious,
however, that it could not have been effective enough to counteract
other forces which prevailed as well.
111. Disturbance of the airflow over part of the starboard wing -
created by No. 4 engine being in a state of reverse thrust-resulted in
appreciable loss of lift and consequent incurrence of a strong rolling
moment to starboard. As the aircraft flew out of ground effect, induced
drag increased and prevented the aircraft from accelerating so that,
at the low air speed prevailing, insufficient control forces could be
obtained by the pilot to prevent the starboard wing from dropping and
to counteract all yaw. Angle of bank therefore increased. With this,
angle of sideslip also increased and decreased angle of attack and further
loss of lift resulted. Height loss culminated in the starboard wing tip
striking the ground. Obviously, circumstances might have been vastly
different and the pilot provided with sufficient control forces to counter-
act all rolling and yawing moments had a higher air speed been reached.
Under the prevailing circumstances, V,,, was of very critical importance.
T h e Importance of V,,,,,
112. With Nos. 1, 2, and 3 engines delivering take-off thrust and
with No. 4 in forward idle, V,,, of ZK-NZB in wings-level attitude
would have been 113 knots. And since V,,, of any aircraft must
necessarily be lower than V, (in this case 118 knots), the pilot would
have been able, had his aircraft been flying in this specific engine
configuration, to keep it under full control and make a climb-out on
three engines.

113. With Nos. 1, 2, and 3 engines delivering take-off thriist and with
No. 4 in reverse idle, V,,, of ZK-NZB in wings-level attitude would have
been 141 knots. The highest air speed achieved by ZK-NZB after
rotation was, however, 16.5 knots short 01 ,that particular requirement.
Eut since, very soon after rotation, the aircra’ft began to roll to starboard
and incurred unfavourable bank (;.e., bank toward the failed engine),
the minimum speed at which it could have been controlled would have
increased over and above the basic V,,, by 3 knots for every degree of
bank attained. The plain fact of the matter is, therefore, that the
aircraft never reached a speed at which it could be properly controlled.

114. When this fact is appreciated, any discussion in respect of what

positive action the pilot could or might have been able to take to rectify
a situation which became critical as soon as the aircraft began to roll
to starboard becomes pointless.

Previous Incident
115. In para. 80 of this report, mention has been made of a previous
incident in which reverse thrust had been unintentionally incurred after
a touch-and-go landing on a training flight. The pilot had managed to
keep the aircraft approximately level laterally but had been unable to
counteract all yaw. I n that case, air speed was reported to have been
130-140 knots but the reverse thrust condition had been recognised
and eliminated quickly.

116. It will be recalled that in that case Captain McLachlan had

moved No. 4 power lever rearward with his thumb and index finger on
the knob. Tests have shown that if a “follow-through” movement is made
with the fingers in that position, it is possible to produce an upward
movement of the thrust brake lever which may result in the latter
moving into the reverse idle detent. The possibility must remain that
on that earlier occasion Captain McLachlan believed that he had
himself induced movement of the thrust brake lever and that he was
determined thereafter to guard against repetition. Thus, he may well
have decided to use the spoiler disarm extension, rather than the knob,
of No. 4 power lever to achieve that purpose. This, however, must
necessarily remain conjectural. Whatever the reason for the incurrence
of reverse thrust on that occasion may have been, it must be concluded
that Captain McLachlan could not have realised that inertia forces
incurred through a very rapid rearward movement of a power lever
might raise the thrust brake lever into the reverse idle detent, otherwise
he would have been expected to report what had occurred.
c

.
Removal of Reverse Thrust from No. 4 Engine
117. It could not be established for how long the reverse thrust configu-
ration had persisted after lift-off. Behaviour of the aircraft, however, was
indicative of persistence at least until the angle of bank had become
appreciable, for had reverse thrust been recognised and eliminated when
the aircraft was banked at a very shallow angle and before sideslip of
any magnitude had developed, acceleration provided by Nos. 1, 2, and 3
engines in take-off thrust and by a gradual increase in thrust from NO. 4
during spool-up might well have allowed the aircraft to be controlled
before it came down to ground level. The condition in which No. 4
engine was found showed beyond doubt that the fan cascade doors had
closed, the primary cascade sleeve had moved forward, and the reverser
buckets had become deactivated before the engine struck the ground.
Thus, the condition of reverse thrust had been recognised on the flight
deck and eliminated. It is clear, however, that remedial action had not
been taken early enough to prevent the aircraft from striking the ground.

118. If it should be claimed that Captain McLachlan ought to have

seen the reverse indicator lights illuminated - there is no evidence to
prove conclusively that they did come on-it could be said, first, that
he would not have been expecting a reversal indication with simulated
engine failure and would probably have been looking outside to observe
closely how First Officer Ruffell was coping with yaw during a one-
engine-out exercise. Again, since First Officer Ruffell was flying the
aircraft, Captain McLachlan would have been unable actually to feel
the yawing forces produced because his feet would have been only
very lightly touching the rudder pedals or in such close proximity to
them that he could have taken over control at any instant. That he
must have seen the reverser lights illuminated or, conversely, have
noticed No. 4 thrust brake lever in the reverse idle detent at some
moment before the aircraft struck, is apparent from the post-accident
state of the reversal components which clearly indicated elimination
of reverse thrust.

119. No. 4 EPR gauge was found to be indicating 096, and since
zero thrugt is represented by 1.0, it is quite clear that the engine was
producing no forward thrust when it separated from its pylon.
The indicator pointer and digital counter tend to remain in their last
indicating position if the system electrical power source is shut off
(vide Systems Operation Manual) and the readings of all four gauges
are therefore of some interest. EPRs for Nos. 1, 2, and 3 engines- 1.24,
1.1 1, and 1.07 respectively - are appreciably lower than the common
EPR set-up for all engines on take-off and suggest a possibillity that all
engines may have been cut just before, at, or just after wing tip impact
and had spooled down to achieve the ratios indicated when the electrical
supply source was disrupted. If power had so been cut, it is possible
that EPR of No. 4, following elimination of reverse thrust, was in excess
of the indicated 0.96 and may have been greater than zero thrust.

120. First Officer Ruffell had not felt the stick shaker operate and
since the aircraft was to all intents and purposes flying very close to the
stall this mizht be regarded as unusual. Disturbance of the airflow in
the vicinity of the sensor (caused by No. 4 engine being in reverse
thrust) , however, could have interfered with the sensor's ability correctly
to monitor angle of attack.

CONCLUSIONS

121. (1) All members of the active flight crew were properly licensed,
medically fit, and qualified to undertake the duties in which they were
currently engaged.
(2) The aircraft was covered by a valid Certificate of Airworthiness
and Maintenance Release and had been properly maintained.
(3) No failure or defect was present or occurred in the airframe,
flying controls and allied systems, or instrumentation before the aircraft
struck the ground.
(4) Until they became dislodged at im?act, all four engines were
responding faithfully to commands transmitted to them through their
respective controls.
(5) Simulated failure of Yo. 4 engine was initiated at or immediately
after VI by Captain McLachlan who moved the appropriate power lever
rearward very rapidly.
(6) The manner in which that power lever was handled and the
rapidity wirh dhich it was moved did not contravene any operating
instruction.
(7) Inertia force produced by the rapid movement of the power lever
caused No. 4 thrust brake lever to rise and enter its reverse idle detent.
(8) After a previous incident in which reverse thrust had unintention-
ally been incurred, Captain McLachlan had not been aware that inertia
force alone might on that occasion have caused the thrust brake lever
to enter the reverse idle detent.
(9) Entry of the thrust brake lever into the reverse idle detent during
the last take-off roll activated the reversal mechanism which caused
the engine to exert powerful reverse thrust during the initial stages of
spooldown, particularly at and immediately after rotation.
(10) First Officer Ruffell was able to maintain directional control
while the aircraft was rolling on the runway largely because the nose
wheel was still in contact with the ground.
(11) The fan cascade doors of No: 4 engine opened fully between VI
and V,.
(12) A state of reverse thrust was not detected by any occupant of
the aircraft before it became airborne.
(13)The aircraft was rotated at the predetermined V, speed which,
however, was much lower than that necessary to allow the pilot to
control the aircraft in the thrust imbalance configuration incurred.
(14) Though rotation was effected more quickly and the nose initially
assumed a steeper pitch-up than usual, neither factor contributed toward
the occurrence of the accident.
(15) When the aircraft rotated, and for an undetermined time there-
after, No. 4 engine remained armed for reverse thrust.
 (16) The amount of rudder applied after lift-off w a s effective in
preventing a large amount of yaw.
(17) Thrust imbalance coupled with loss of lift occasioned by disturb-
ance of the airflow produced by No. 4 engine in reverse thrust, induced
a strong rolling moment to starboard together with sideslip, which
persisted until the wing tip struck the ground.
(18) After lift-off, the aircraft did not attain sufficient air speed to
enable the pilot to obtain full control over its movements.
(19) Failure to accelerate and to gain height were due to increase of
induced drag and loss of lift resulting from uncontrollable roll effects.
(20) At an undetermined time after lift-off, reverse thrust was recog-
nised and eliminated.
(21) No. 4 engine was delivering no thrust when it became separated
from its pylon.
(22) After elimination of reverse thrust, insufficient time and height
were available to enable recovery to be made before the wing tip struck.
( 23) Neither weather consdiitions, runway conditions, nor communica-
tions services were factors contributing toward Occurrence of the accident.
(24) If a power lever of a DC-8 aircraft is moved rearward very
rapidly, it is possible for inertia force thereby generated to cause the
associated thrust brake lever to rise and enter the reverse idle detent
and so a m the engine for reverse thrust.
(25) I n a previous instance in which reverse thrust had been incurred
unintentionally, no occupant of the aircraft involved had been aware
that that condition might have been caused by inertia force produced
when the power lever was moved rearward.
(26) Prior to the occurrence of the accident under review, neither
Air New Zealand Ltd. nor the Department of Civil Aviation in New
Zealand was aware from its own experience or h?d learned from any
source elsewhere that a state of reverse thrust might result from very
rapid rearward movement of a DC-8 power lever or of a power lever
of similar design and mode of operation used in other types of aircraft.

OPINION
122. T h e primary cause of this accident was the incurrence of reverse
thrust during simulated failure of No. 4 engine on take-off.
That condition arose when very rapid rearward movement of the
power lever generated an inertia force which caused the associated
thrust brake lever to rise and enter the reverse idle detent.
After lift-off , the minimum control speed essentially required to over-
come the prevailing state of thrust imbalance was never attained and an
uncontrollable roll, accompanied by some degree of yaw and sideslip
in the same direction, ensued.
When the condition of reverse thrust was recognised and eliminated,
insufficient time and height were available to allow the aircraft to
recover from its precarious attitude before it struck the ground.
 COMMENT
Handling of Power Levers
123. In earlier days of flying, many a n insltructor made a habit of
closing a throttle suddenly to create an element of surprise in the
trainee and to determine how he would react to the equivalent of a n
unexpected genuine emergency. In the piston engine era, sudden closure
of a throttle (inviting or resulting, in some cases, in unwanted conse-
quences) produced an immediate loss of engine power and propeller
thrust.

124. With the turbojet engine, the rapidity with which a power lever
is “closed” does not result in a reduction of thrust at a rate proportional
to rapidity of power lever movement. Operation of the power lever
activates a fuel flow scheduling system which, for very important reasons,
limits thrust decay to a rate that cannot be exceeded no matter how
quickly the pilot may move the lever back. The system “takes its own
good time” and this incurs no disadvantage. It may be argued, then,
that since, for instance, a Pratt and Whitney JT311 engine takes seme
14 seconds to spool down from take-off thrust to forward idle, there is
neither merit nor advantage to be gained from moving its power lever
very rapidly rearward.

125. On the contrary, it may be argued that though an instructor

must well know that turbojet engine deceleration cannot be hastened
beyond a fLved rate, he is equally aware that a power lever may be
operated very rapidly without incurring the slightest xisk of engine
damage, flame out, or other unwanted circumstance. If an exercise
involving simulated engine failure on take-off is to be valuable as
practice for meeting an actual failure, power loss must preferably be
simulated at the most disadvantageous moment, i.e., between Vl and V,,
points separated by a bare three seconds of time. Since accurate timing
is important and attention to other matters must be given, the instructor
must act promptly and may in consequence be inclined tcj snap the
power lever back very quickly, thereafter devoting his attention to the
runway ahead to ensure that the trainee maintains directional control
when and after power is pulled off.

126, There is an important psychological factor - possibly the most

easily understood - which has undoubtedly influenced some instructors
to close throttles or power levers suddenly. The more quickly and unob-
trusively a power lever can be moved, :the less time and opportunity a
trainee will have to observe, out of the corner of his eye, which lever
is being handled. If a trainee is able to see which lever is to be moved,
he becomes forewarned of what he will be required to do to cope with
the specific asymmetric thrust configuration intended by the instructor.
The value of the exercise to which he is being subjected may thus be
much reduced.

127. I t is considered that it would be unjust to level criticism a t

Captain McLachlan for the manner in which he rnoved the power
lever. He merely repeated what many other instructors had been doing
hitherto and what happened was totally unexpected and not imme-
diately recognised. If, however, the demonstrated consequences of very
rapid rearward movement of a DC-8 power lever are now apparent and
may be held to be wholly unacceptable, a remedy is offered in comment
which follows.

Prevention of Recurrence
128. Investigation of this accident and trials conducted with several
other DC-8 aircraft have demonstrated clearly that in some cases inertia
force generated by very rapid rearward movement of a power lever
will actually cause the associated thrust brake lever to rise and enter
the reverse idle detent. Though the possibility of a recurrence of the
circumstances might be very considerably reduced by adopting a special
technique for handling the power lever and by providing an appropriate
monitoring method, the condition might still occur some time. For it is
a well recognised fact that if a particular thing can be done, albeit quite
unintentionally, then sooner or later some person will do it. If, therefore,
it ought properly to be the aim to make repetition of the circumstances
of this accident totally impossible, some form of mechanical protection
against unintended movement of the thrust brake lever must be incor-
porated into the system. Whatever may be done should afford that
protection without making power lever movement any more difficult
than it now is and without adversely affecting simplicity of operation
and time taken to move the thrust brake lever when required. No
alternative is worth considering if the remotest chance exists that some
unwanted consequence wiIl arise from its introduction. A specific
Recommendatim is made in para. 134 of this report.

Training

129. Some training exercises routinely carried out in modern swept-

wing jet transport aircraft have resulted in accidents. In examining
aspects of modern flight crew training it is appropriate to consider
whether some exercises ought to be modified in the light of recent
experience in different parts of the world. I t has been suggested by some,
principally by thoughtful laymen or by others not intimately familiar
with operational practice, that certain exercises ought to be abandoned
entirely in that the type of emergency being practised is never likely to
occur in airline operations.
130. I t would be unrealistic to expect one airline to make radical
changes in its own training syllabus without the support of the majority
of others but it does appear from a study of training accidents that have
occurred in reccnt times that a thorough review of training exercises
and procedures on an international level would be profitable. In this
regard ICAO, preferably in close collaboration with IFALPA, would
appear to be the organisation best fitted to look into the matter.
Modification, rather than complete abandonment, of certain exercises
appears desirable, for since responsibility for the lives of many passengers
and the preservation intact of a valuable piece of equipment is of
primary importance, no pilot should be afforded an opportunity of
becoming complacent through some decision that, because he will be
unlikely to encounter a particular emergency but will probably be able
to C ~ J I Jwith
~ it if it should occur, there is little point in providing him
with practical training to meet it.
131. Though training on flight simulators as an alternative to using
aircraft for many exercises involving emergency procedures has be:n
suggested by some without a full appreciation of what is involved UI
operational flying, it is necessary to mention that while simulators are
valuable aids toward obtaining proficiency in certain procedures, they
can by no means entirely replace aircraft and the type of experience that
the latter afford. In any event, the circumstances of the accident under
rfview have not shown that the accident would not have occufred had
smulators been used to supplement the practical flying trainmq. .The
question need not be further dealt with here, but if or when tra1nm.g
policy is discussed at international level, it is important that each and
every conclusion reached and resolution adopted be conditioned by a
full appreciation of the fact that the fare-paying public has a rlght
to demand that crews of aircraft have been proven capable of handllng
satisfactorily any emergency, no matter how remote the likelihood of its
actual occurrence may appear to be.

Sharing of Experience
132. “Share your experience” is a familiar, widely accepted and very
important flight safety precept. A possible explanation for Captain
McLachlan’s omission to report a previous incident in which reverse
thrust was unintentionally incurred has already been offered. I t must
be regarded as certain that had he realised that reverse thrust might
have been caused by inertia force generated by hand movement he would
have reported the incident without hesitation. That in which he was
involved prompts consideration of the question as, to what kinds of
experience, and how much, ought to be shared. To this there can be
but one answer, for it is based on a great many lessons learned from
the past: Share every experience, report every incident immediately,
however trivial it may be considered to be.

Emergency Services
1.33. An inquiry was held by the Department of Civil Aviation to
review all aspects of rescue, fire fighting, and crowd control measures
taken in the case of this accident and to determine what lessons had
been learned from its occurrence. As a result of prompt steps taken to
rescue the occupants of the aircraft, no further loss of life or personal
injury resulted.

RECOMMENDATION

134.I n order to prevent a recurrence of an accident of this nature,

it is recommended :
( 1 ) That engine handling technique be revised to ensure that the
rate at which any power lever is moved back is insufficient to create
inertia to cause its associated thrust brake lever to rise toward or enter
on
the reverse idle detent; additionally, that the power lever always be
held in such a manner that rearward movement of the hand does not
impart a rotary motion to :the thrust brake lever.
(2) That some form of mechanical protection be incorporated in the
power lever and associated thrust brake lever system so that, should the
measures recommended in (1) above not be adhered to, incurrence of
unwanted reverse thrust wiIl be rendered impossible.

FURTHER INQUIRY

135. The cause of this accident appears to have been reliably estab-
lished and it is therefore considered that further inquiry into the
circumstances would serve no useful purpose.

COMPLIANCE WITH REGULATIONS

136. Since it does not appear that any ‘degree of responsibility for
the occurrence of this accident may fairly be attributed to any person
or organisation, the provisions of regulation 10 (3) of the Civil Aviation
(Investigation of AccideEts) Regulations 1953 are deemed inapplicable.

0. J. OBRIEN,
Chief Inspector of Accidents.
E. F. HARVIE,
Inspector of Accidents.
10 September 1966.