Voice recorder and Accident

4

Cockpit voice recorder

In accordance with national and international requirements, TC-JAV was equipped with a cockpit voice recorder (CVR), type Collins, model 642 C-1. The cockpit voice recorder was located alongside the flight data recorder, inside the aft cargo compartment and immediately to the rear of the aft cargo door on the left and side.

The CVR was found about l50 metres beyond the initial point of impact in the forest of Ermenonville and was considerably damaged, but the recording could still be heard. It was first played beck and copies made at the French Air Navigation Technical Service, in the presence of French and foreign experts of the Commission of Inquiry.

Appended is the transcript of this recording; the times were established in agreement with the time tracks of the communications between the aircraft and the central services (Airport and ACC North).

'It should be noted that the time which elapsed, seventy«seven seconds, between tho noise of decompression and the end of the CVR recording is practically identical with that measured on the flight data recorder. Nevertheless, for the period in question there is a time shift of the order of thirty seconds between the time supplied by the flight data recorder (indication taken from the clock of the flight engineer’s pane1) and the time common to the ACC/Tower/CVR (ACC time = flight data recorder time minus 30 seconds). With allowance for the various factors available for the purposes of comparison, this discrepancy appears to be approximately constant and of the same magnitude throughout
the flight.

In addition to the information provided by the communications recorde4 on the ground and the various observations entered on the transcript appended, the CVR recording reveals in particular the following points during the 1ast phase of the fight:-

Decompression heard at 113956 hrs.

Pressurization warning heard almost immediately and for slightly less than 25 seconds.

Identification of the nature of the accident by the crew.

Overspeed warning heard at about 114023 hrs, probably untii the end of the record-ing (although very faint during the last moments) ie for 50 seconds.

CVR recording stopped at 114113 hrs.

Flight data recording

In accordance to the national and international requirements, TC-JAV was equipped with a flight data recorder, type Sunstrand Data Control, model 573 A, maker’s serial number 2104, located under the floor on the left-hand side, alongside the cockpit voice recorder and immediate1y to the rear of the aft cargo door on the left-hand side.

The fight data recorder was found in the area of the main wreckage, about 60D m 5om the initial point of impact. The outer case was substantially damaged on impact (no trace of fire or smoke); the vicaUoy magnetic tape was very dirty, bent and broken in two places. The recorder was taken to the USA where the NTSB arranged for the tape to be read out at the premises of the firms of Sunstrand and Teledyne, in the presence of French experts from Bretigny Flight Test Centre and the 'Accidents Investigation Bureau. The read-out was established in graphical form and subsequently a fresh read-out was made at the Brbtigny Flight Test Centre on the RESEDA installation, using the calibrations established for the KSSU group. The results obtained were in agreement.

Appended are graphs showing the variations in the parameters during the flight and the accident.

The principal points regarding the read-out may be surmised as follows:

Take-off: 1131 hrs (flight data recorder time).

Takeoff was made with reduced thrust, rotation began at 143 knots, ie 4 knots below the predetermined V2, and the ground roB time was of the order of 40 seconds.

During the takeoff, the horizontal stabilizer setting was about -6%º, the maximum attitude reached was 19', while the elevator angle changed progressively from 8º to l 1º.

Climb: 1132 hrs to 1.139 hrs (flight data recorder time).

The climb progressed normally. It should be noted that there was a stretch of level flight at 6,000 feet for more than two minutes (l135 hrs to 1136 hrs). Shortly before reaching 12,000 feet, the aircraft climbed at 300 knots indicated air speed with a rate of climb of 2,200 feet a minute. The attitude was of the order of 3', the horizontal stabilizer was set at ¿A' nose-up and the elevators were 2º to 3º aircraft nose-up.

Accident:

It can be deduced that pressurization occurred at 114026 hrs at about 11,500 feet. Two seconds later the following conditions were found:

An angle of some 10º, to the left of the two rudder control surfaces; the change in heading to the left was 9º.

A nose-down movement of the elevator surfaces of which the angle had decreased by about 3º, while the aircraft’s attitude decreased accordingly.

The horizontal stabilizer angle recorded changed from ’hº nose-up to 6Viº nose-down.

The speed of No 2 engine had fallen to 459o by 114029 hrs.

(Only one item of this information is aberrant, that relating to the setting of the horizontal stabilizer, limited by a mechanical stop to an angle smaller than that recorded after the depressurization. In addition, the horizontal stabilizer cannot move so swiftly and a change in angle as large as that recorded would inevitably have given rise to far larger vertical accelerations than those recorded. The transducer of the movement of the stabilizer is in a forward position in the aircraft, on the cable which transmits this information to the cockpit, and it is more than likely that the doubtful value recorded corresponds to tension of this cable, associated with the damage to the floor.)

The aircraft nose-down attitude became rapidly steeper, -20º was reached 22 seconds after decompression and simultaneously the speed increased to 362 knots, although Nos 1 and 3 engines had been throttled back.

At the end of minute 1140 hrs, the speed reached 400 knots at 7,200 feet, the ailerons seemed to function correctly and the angle of bank to the left did not exceed 20º.

At the beginning of minute 1141 hrs, the attitude began to decrease progressively and the speed to stabilise around 430 knots without this appearing to be due to the control surfaces.

Impact occurred at 114143 hrs (flight data recorder time) at an attitude of -4º, a speed of 423 knots and on a heading of 281º.

Wreckage
Main wreckage

The DC-10 made impact with the ground in the department of Oise at a place called ‘Le Bosquet de Dammartin’ in the commune of Fontaine ChaaUs.

The accident site is located in a small enclosed valley, running from east to west and covered with Scotch and maritime pines. The ground is rugged with some rock outcrops on the east side. The average elevation is 105 metres. The area affected by the aircraft impact exceeded 65,000 square metres. The aircraft literally disintegrated into fragmented wreckage. It cut a swath through the forest some 700 metres long by 100 metres wide.

On the initial impact with the tops of trees about 10 metres high, the aircraft was on a heading of 280º, with a pitch attitude close to -4º and an angle of bank to the left of the order of 17º. This information has been obtained from the flight data recording and inspection of the site, both of which gave the same indications. The very high speed was between 420 and 430 knots (800 km/hr).

If this initial impact on the edge of the ‘des Epines’ forest road is taken as the point of origin for the measurement of distance on the wreckage trail, it becomes evident that in the violence of the impact the airframe was completely shattered. The wreckage recovered came from both forward and rear parts of the aircraft structure.

Impact with the gr6und was made about 330 metres away from the initial 'point of impact. Along this distance of 330 metres the aircraft cut through hundreds of trees. Wreckage of the wings and the frames of the forward doors was found along a trail 100 to 150 meties wide. At a point about 220 metres along and 60 metres south of the centre line of the swath, the cockpit voice recorder was found it had been installed in the rear part of the aircraft on the left-hand side, below the leading edge of the vertical stabilizer.

Between 170 and 270 metres along the wreckage trail, traces of kerosene were found along both the north and south edges of the va11ey, providing evidence of 'the points where the fuel tanks broke up. Wreckage of No 3 engine was also found on the north side of the area devastated by the aircraft

Between 250 and ‘270 metres, along this same north side, numerous small pieces of wreck-age were found, mainly of the wings and the engines. The airframe touched the ground at a point between 330 and 440 metres along the trail. A violent explosion ensued and the practical disintegration of the aircraft. In this area, numerous small fragments from all parts of the aircraft were found intermingled.

From 400 to 600 metres, various pieces of wreckage, generally small in size, were found scatted over the whole width of' the crash area. The flight data recorder was found at the left-hand edge of this ‘section of the trai1. At the end of this section there were two fairly large pieces of wreckage, the tail aft body and part of the fuselage with a door frame and nine windows.

No 2 engine was recovered at 650 m, on the ‘de la cavee road. This engine had remained sufficiently intact for its component parts to be examined.

The last wreckage eas found 700 metres fiom the initial point of impact.

Wreckage found at Saint-Pathus

On the morning of 4 March, French experts accompanied by police officers from Saint-Pathus found the bodies of six passengers, parts of the aircraft seats and the wreckage of the aft cargo door, beneath the aircraft’s flight path and 15 km before the main wreckage.

Aft cargo door on the left-hand side

The wreckage of this door was composed of:

The lower part of the door with the 4 latches and their complete control and locking mechanism, and a1so the push rod controlling the lock tube.

The electric motor of the latch actuator was not found.

Part of the door including the locking handle with the link and the vent door shaft and also the detached tap fitting of the push rod.

The vent door was not found. The cargo door operating arm.

All these parts came down in free fall into newly ploughed fields and were embedded in the fairly soft earth so that they were little damaged on impact with the ground; any fractures occurred on ejection from the aircraft.

A visual inspection of the wreckage was carried out immediately at the sites where it was found, with the following results:

absence of any marks of fire or over-heating,

incomplete closing of the door latches;

non-engagement of the lock-pins;

the electric motor of the latch actuator had become detached from its mounting and was not found;

the handle was out of its housing (open position); the trigger whicb retains. it in its housing had been forced and was difficult to work;

the links controlling the latch had not reached over-centre and any slight force exerted on the latches caused displacement of the actuator assembly of which the upper part was no longer fixed positively to the door structure.

All these parts were then taken to Le Bourget and subsequently to Saclay Tqst Centre for more detailed laboratory examination.

Identification of the door

Two references were ink-stamped on the door in two different places:

First reference: P/N NFA 6070-501 N

S/N 46704/11 FG401
Second reference: P/N NFA 6070-507
The following indications were inscribed under the reference:
F/N EDITION PROD INSP ACCEPT

29 1 058D4 ZAI 04

The document DOUGLAS NFA 6070 ‘W’ confirmed the door reference of NFA 6070-507 (the vent door installation had caused the reference to be changed from 501 to 507).

Principal findings of the examination of the door wreckage

On the instruction plate for manual opening of the latch actuator, there was a hole, 28 not prescribed by. Douglas and driRed by THY, in order to gain direct access to the drive mechanism (incorrect exeeution of SB 52-3S).

The link (LINK ASSY P/N ADA 7366-501), between the locking handle and the vent door. shaft, was bent

The additional support plate specified for the vent door shaft by SB 52-37: had not been installed‘.

The push rod (LINK ASSY P/N ADA 7372), between the vent door shaft and the lock tube, was. beat and the two crank attachment rivets were sheared. ‘

The forward bottom structural corner of the door was deformed. This damage, due to contact with the ground after a fall of 3,600 metres, had caused slight deformation of the Iock tube and the mounting of the lock limit switch.

The end of the lock tube was chamfered as prescribed by SS 52-37. Rough file marks and irregular scoring showed that this work was done man

The stricker of-the unlock switch had two DOUGLAS P/N A'FA 3210-1 shims, surmounted by a third shim with no reference number consisting thin crumpled piece of metal leaf with numerous folds on the side on the side which had to come up against the roller of the unlock limit switch. The presence of this part, unusual in equipment to aeronautical standards, was surprising and could only have entailed imprecision and erratic functioning of this switch which closes the circuit to the latch actuator in the sense of opening the latches only. This defective installation had no effect on safety but could have been the source of numerous difficulties in opening the door.

Medica1 and pathological information

It had already been decided on the day after the accident to take the remains of the ’ passengers and crew members to the Institute Miedico-Legal de Paris, for the purpose of. the Inquiry.

In view of the exceptionally large number of victims t4e medical team encountered difficulties, as the institute did not have facilities on a scale related to this type of accident.

The results of the examination of the bodies of the victims were as follows:

Lesions observed

From the traumatological standpoint, the lesions could be classified into two categories:

(a) In the case of the bodies recovered at the main accident site in the forest of Ermenonville, there was a high degree of fragmentation (nearly 20,000 fragments were listed) associated with the violence of the impact.

(b) On the other hand, the six bodies found near Saint-Pathus were complete, although presenting fractures and serious visceral lesions. Careful examination showed:

that there were no external burns;

that there were no external lesions which could be associated with the projection of metal or other fragments as the result of an explosion caused either by a criminal act or otherwise;

by X-ray examination, that there was no evidence of deep penetration by metal fragments.

Toxicological examination

Routine toxicological examinations were made of the six bodies recovered over Saint-Pathus and of a number of fragments recovered at Ermenonville and selected at random. No evidence was detected of:

carbon monoxide;

cyanhydric derivatives;

chlorine derivatives;

alcohol.

Identification

188 bodies or parts of bodies were positively identified by the use of a number of techniques:

finger-printing (in particular in the case of Turkish and Japanese nationals, because of the existence of national finger-print records);

examination of teeth;

bone measurements;

clothing, personal effects.

It should be noted that finger-printing was of' great assistance and that the use of a computer proved to be essential for processing the enormous quantity of data required for identification purposes.

Fire:

The circumstances of the impact (disintegration at very high speed in the trees) were such that there was practically no fire, apart from a few very minor localised outbreaks where fires of short duration persisted for only a few moments.

Jet Al fuel was used and there were about 23,500 Btres on board at the time of impact.

Survival aspects – Rescue operations

Survival aspects

The accident occurred in two phases which left no chance of survival for the 346 occupants of the DC-10. The first phase occurred at 1140 hrs over the commune of Saint-Pathus, where six of the aircraft occupants were ejected at an altitude of about 3,600 metres.

The second phase was that of the impact at 1141 hrs, when the aircraft flew into the forest at a speed of 430 knots (800 km/hr), leaving no chanci of survival for its occupants.

Rescue operations

The Air Traffic Control was immediately aware of the loss of radio and radar contact and was able to locate the area of the accident, thereby simplifying the task of the alerting and search services (VHF caBs, to which no reply was received, from 114150 hrs to 114650 hrs; telephone enquiries in conjunction wiih Orly, Le Bourget and Creil, from 1144 hrs to 1238 hrs).

The rescue operation was already well under way when the DETRESFA message was transmitted by the North Area Control Centre at 1200 hrs and the presumption of accident message at 1240 lus.

After information had been obtained at the accident site, the accident notification message was transmitted at 1615 hrs and the DETRESFA termination message at 1650 hrs.

At 1840 hrs, Doullens Centre announced that the SAR operation had ended at 1732 hrs.

In addition to the occurrences recorded by the North Area Control Centre, the crash of the DC-10 was reported by 1145 hrs to the Senlis Gendarmerie Brigade by the Survilliers C.R.S. (riot police) Station (Val d’Oise). From 1145 hrs onwards, exceptionally large scale rescue facilities (air and ground) were put into operation by the civil and military authorities (police, armed forces, civil emergency centres, Paris Airport, etc).

Within a very short time, at 1215 hrs, the first rescue teams arrived at the site. The conveyance of the bodies of the victims to the church of Saint-Pierre de Senlis began at 1345 hrs.

The bodies recovered near the villages of Saint-Pathus and Oissery were taken to Meaux hospital.

Seventeen emergency centres (civil and military facilities) with fifty-six vehicles of various kinds were used and about three hundred persons took part in the operations on the first day.

Finally, the operations for the transfer of the aircraft wreckage began on 8 March and were completed on 20 March.