Middle East Airlines SE 210 Caravelle III OD-AEM accident at sea 10 NM SSE of Dhahran Airport, Saudi Arabia on 17 April 1964.

Report dated July 1964 of the Committee of Accident Investigation convened by the Superintendent Director General of Civil Aviation, Saudi Arabia

Prepared by Harro Ranter

Source: Aircraft Accident Digest No.16 (ICAO Circular 82-AN/69) page 151-161.


1.1 History of the flight
1.2 Injuries to persons
1.3 Damage to aircraft
1.4 Other damage
1.5 Crew information
1.6 Aircraft information
1.7 Meteorological information
1.8 Aids to navigation
1.9 Communications
1.10 Aerodrome and around facilities
1.11 Flight recorders
1.12 Wreckage
1.13 Fire
1.14 Survival aspects
1.15 Tests and research
2. Analysis and Conclusions
2.1 Analysis
2.2 Conclusions
3. Recommendation

1. - Investigation

1.1 History of the flight

Flight ME 444 was a scheduled international flight from Beirut to Dhahran Airport. It departed Beirut at 1709 hours GMT and proceeded in accordance with its flight plan to Dhahran at flight level 300. At 1904 hours the aircraft reported to Bahrain Control that it was estimating Dhahran at 1928 hours, and was cleared to descend to reach flight level 50 over the Dhahran beacon. At 1906 hours it contacted Dhahran Tower and requested the latest wind and visibility, which were given as NNE/10 kt, gusting to 16,and 1/2 NM (reported 110 yd in suspended dust). At 1909 hours the flight reported to Bahrain that it was leaving FL 300, and at 1926 hours that it was estimating the Dhahran NDB in two minutes. At 1928 hours it contacted Dhahran and reported "5 000 ft descending". It was cleared for an ADF approach and requested to report at 4 000 ft and outbound at 2 000 ft, QNH 1 006 mb. At 1929 hours it reported leaving 4 000 ft and at 1930 hours passing 2 500 ft and turning inbound. It was then cleared to final approach and requested to report reaching minimum and runway in sight. At approximately 1932 hours a short loud transmission noise was recorded by the Tower. No further message was received from the flight. It was subsequently found that the aircraft struck the sea at the completion of the procedure turn 4 NM off shore and 10 NM south of Dhahran Airport (26�'55"N - 50�'36"E). The accident occurred at 1932 hours GMT.

1.2 Injuries to persons
Injuries Crew Passengers Others
Fatal 7 42-
Non-Fatal - --
None - --

1.3 Damage to aircraft

The aircraft was destroyed.

1.4 Other damage


1.5 Crew information

The pilot-in-command, aged 33, held a valid airline transport pilot's licence with a Group 1 rating for Caravelle aircraft. He had flown a total of 9 193 hours, including 3 425 hours as pilot-in-command and 235 hours on Caravelles, of which 10:35 were in OD-AEM *1)

The co-pilot, aged 36, also held a valid airline transport pilot's licence with a Group 2 rating for Caravelle aircraft. He had flown a total of 7 691 hours, including 1 680 hours as pilot-in-cormnand and 70 hours on Caravelles, of which 29 hours were in OD-AEM *1)

The flight engineer, aged 42, held Lebanese and French flight engineer licencrs with a Class I rating for the Caravelle SE 210. He had flown a total of 15 000 hours, including more than 1 500 hours on Caravelles, mostly on Caravelle III aircraft.

*1) OD-AEM was the only Caravelle III of the MEA fleet. The pilot's instrument panel of this aircraft differed from the instrument panel of Caravelle VI N aircraft as follows:

Instrument Caravelle III Caravelle VI N
Altimeter Kollsman single pointer drum Smiths two pointers
Airspeed Indicator Smiths two pointers Kollsman single pointer (100's) drum (10's)
Radio Altimeter Yes None
RMI Separated by radio altimeter Adjacent
Director Horizon Different presentation
Standby Horizon Below radio altimeter Below stop-watch
Turn & Bank Indicator Below ILS indicator Below VOR RMI

The co-pilot's instrument panel of this aircraft also differed in many instances from the instrument panel of Caravelles VI N.

1.6 Aircraft information
The aircraft was a Caravelle III. It had a valid Lebanese certificate of registration and a certificate of airworthiness valid until 29 January 1965. A certificate of maintenance, valid for 350 hours, had been issued for the aircraft on 5 april 1964.

At the completion of a flight Beirut-Ankara-Beirut on that same day some technic~ defects were reported and were corrected prior to the departure of the aircraft for Dhahran

The maximum gross weight allowed for this flight was 46 000 kg. At the time of the accident the aircraft weight was estimated as being approximately 37 250 kg. At the commencement of the flight the centre of gravity was at 32.2%, well within the limits (25%- 39% MAC).

1.7 Meteorological information

The forecast for Dhahran supplied to the flight at 1025 hours at Beirut was as follows:

Valid from 18 - 2400 hours -
visibility 15 NM, intermittently 2 NM falling to 01 NM in sandstorm; wind, 140�/15 kt, gusting 25 kt; intermittently, 340�/25 kt, gusting 40 kt; cloud 4/8 Sc 3 500 ft, 5/8 Ac 15 000 ft; temporarily 1/8 Cb 3 500 ft.

At the time of the accident the weather conditions at Dhahran Airport were:

cloud ceiling, sky obscured; visibility 110yd in dust haze; wind 10-20�/16 kt, gusting 22 kt; temperature 28,5癈; dew point 8,5癈.

1.8 Aids to navigation

Aids fitted to the aircraft were: VOR, ADF, ILS and a radar scope.

The ADF and VOR both appear to have been indicating correctly.

Aids available at Dhahran: VOR situated on the aerodrome about 500 yd east of runway 34/16 and 2 000 yd north of the threshold of runway 34. NDB (DH) situated 1.2 NM SSE of the threshold of runway 34, with which it was aligned.

1.9 Communications

All radio communications between the aircraft and air traffic control at Bahrain and Dhahran were normal.

1.10 Aerodrome and around facilities

Dhahran Airport was fully operational throughout the aircraft's approach. The main runway was in use - 34/16, 10 000 ft by 200 ft. The approach lights were not illuminated as they were under repair.

1.11 Flight recorders

Not mentioned in the report.

1.12 Wreckage

Salvage operations were commenced the day following the accident and about 95% of the aircraft structure was recovered in an area of about 250 ft radius from the main body of the aircraft. At the moment of impact the aircraft was slightly nose-down and banked to the right. The attitude, together with the nature and extent of the airframe damage, appeared consistent with striking the water at approach speed.

The front and rear fuselage sections, although disconnected structurally, were still loosely held by control and electrical cables and were separated during salvage operations.

1.13 Fire

There was no fire.

1.14 Survival aspects

There were no survivors. Seventy passenger seats were recovered out of a total of 80. No evidence was found to indicate that any seat had suffered damage due to displacement by inertia loads during impact. The release of seats from the forward first class compartment was due to the break-up and displacement of the cabin floor as the front fuselage broke up following impact with the water. The seats in the centre part of the fuselage, which remained attached to the left wing, were still securely fixed in position and virtually undamaged. There was evidence that most of the passengers had been strapped in their seats.

1.15 Tests and research

The following equipment was removed from the aircraft and examined by Air France at Orly under the general direction of the Minist鑢e des Travaux Publics et des Transports:
a) pilot's and co-pilot's altimeters, air speed indicators and vertical speed indicators;
b) both gyro units with pilot's and co-pilot's HZ4 horizon indicators;
c) the radio altimeter, indicator and switch unit;
d) the elevator and rudder feel system;
e) the four servodyne units operating the aircraft's flying control surfaces.

It was not possible to calibrate either of the altimeters, the airspeed indica- tors or the vertical speed indicators due to corrosion of the mechanism following their immersion in sea water. Strip examination of the mechanisms revealed that all barometric capsules were serviceable and no evidence of pre-crash failure or defect was found in these instruments. Both altimeters were set to 1 006 mb.

Complete functional testing of the gyro units and computer could not be carried out owing to the effects of salt deposition and corrosion on electrical components. The directional and vertical gyros and the HZ4 indicator units were tested individually and found to be capable of normal operation.

The radio altimeter transmitter and receiver could not be bench checked due to damage and corrosion. The selector vas at 400 ft and functionad correctly at this position. The indicator was selected to "small scale" with the switch in the "on" position.

Examination and checking of the mechanical portion of the rudder and elevator teel systems revealed no evidence of any pre-crash defect or malfunction, and the settinR of the torsion bars of tne springload system were correct. The pitch-corrector actuator was in the" normal" position. Strip examination of the hydraulic components of the svsteir. i.e. the actuators, tlir pressure reducing valve and the on/off selector revealed no evideni:� of any pre-crash defect or malfunction. The pre-crash position of the on/off selector could not be determined.

The four servodyne units operating ailerons, elevator and rudder were bench tested and found to operate normally. No evidence was found of any pre-crash defect or malfunction.

2. Analysis and Conclusions

2.1 Analysis

It was believed that the time at which the loud transmission noise was tape recorded by Dhahran Tower, 64 seconds after the aircraft had informed the Tower that it was "turning inbound", was the time of the accident. Since the aircraft had been cleared by Bahrain control to 5 000 ft over the Dhahran NDB, it was believed that it was over the beacon at about the time it reported "5 000 ft descending". At normal instrument approach speed the aircraft could have descended to the position where the accident took place in the 3 minutes 26 seconds interval between these two transmissions.

As the Jeppesen NDB Instrument Approach Chart was found loose in the cockpit after the accident, it was presumed that the procedure laid down in the chart was being carried out. The Operator drew attention to the fact that the pilot-in-command was very familiar with the "teardrop" pattern approach procedure which was prescribed for Dhahran. The Jeppesen instrument Approach Chart indicates among other things that the procedure turn should be made in level flight or completed at 1600 ft QNH. After the completion of the procedure turn the aircraft may descend to the OCL of 648 ft (MEA's minimum is 700 ft) and continue at this height inbound until it arrives overhead the NDB.

Evidences indicated that at the time the aircraft struck the water it was in an approach configuration (10� flap setting and undercarriage extended). Its speed could not be accurately established, although it was probably in excess of 154 kt and may have been in the range of 170 - 180 kt. Its attitude on impact was slightly nose-down and banked to the right, which indicated that the aircraft was completing or had just completed the procedure turn. At this point the aircraft should have been at an altitude of at least 600 ft. The technical examination of the wreckage failed to produce any evidence of malfunction or failure of the aircraft, its engines or its equipment. No evidence of explosion or bird strike in flight were found and the manner in which the aircraft struck the water supported the conclusion that the aircraft was operating normally at the time of the accident.

In considering whether the accident might have been the result of the pilot-in- command misreading his primary altimeter or being confused over his flight instruments, the Committee appreciated that under the conditions prevailing during an instrument approach and the environment created by night and sandstorm, both pilots would have been concentrating on their flight instruments. As the co-pilot was known for his mental alertness and habit of closely monitoring and commenting upon any variations from the correct conduct of a flight, and the experienced flight engineer had a reputation for closely monitoring the approach phase of a flight, it was difficult to conceive that if the pilot-in-command had permitted the aircraft to descend dangerously low as a result of misreading his altimeter or being confused over his flight instruments, it could have passed unnoticed by the other crew members. However, the Committee examined these possibilities.

Misreading of the Altimeter

The Committee considered the possibility that the pilot-in-command misread his primary altimeter (pressure drum with single pointer type) by 1 000 ft high and then decided to descend to the OCL during the procedure turn.

The Committee noted the similarities between the subject accident and two previou, Caravelle III accidents, one at Augara (Esenboga) on 19 January, 1960,and the other at Rabat on 12 September, 1961. The two aircraft involved in these accidents had similar instrumentation and were carrying out ADF instrument approaches at night. In neither of these accidents was any evidence found of technical failure or malfunction which could have caused the accidents. In both cases the aircraft were flown by experienced pilots, and the investigating authority stated that the possibility of the pilot misreading his altimeter could not be ruled out.

Following the Rabat accident, the top portion of the altimeter's window was blanked off so that only one complete numeral of the drum could be seen at a time. While the Committee accepted that this measure should have made it impossible for a pilot to misread the instrument by 1000 ft high, nevertheless the Committee did not rule out this possibility. It also considered it possible that the crew might have been misled into a misinterpretation of the height by a false indication on the radio altimeter.

Following trials carried out in the Sahara in 1958, Air France issued the following technical instruction:

"1. These radio altimeters (RCA AVQ 6 and CSF AM 210) are frequency modulated. Measurements in the Saharan laboratory and simulated trials on these radio altimeters have shown that.

- The electric fields surrounding the aircraft are of the order of 1 volt/cm in calm weather, but they exceed 150 volts/cm in violent sandstorms.

- Under these conditions, the static discharge wicks can no longer ensure the continuous discharge of the aircraft and series of rapid discharges arise in those areas of the airframe which have a small radius of curvature; the phenomenon is maintained by the continuous electrification produced by the grains of sand and by a state of partial ionization near the disruption points.

- The rate of repetition of the discharges comes within the range of action of the computing circuits of the radio altimeter,so that interference pulses are added to the normal pulses resulting from the beating of the emitted and received waves. For this reason, there is sometimes an indication of height greater than the actual height above the ground.

2. Under these conditions, the radio altimeter must be used only with the greatest caution. The pressure altimeter must be regarded as the basic instrument for measuring the aircraft's altitude and the radio altimeter should be used only as a cross check of this basic information.

3. It is only in extreme cases of very strong sandstorms that the pressure alti- meter may also give erroneous indications (sand in the intakes etc.). Under these rare and special conditions:

- The indications of the instruments must be cross checked and the instrument giving the lowest altitude must be used.

- The approach must not be continued below 500 ft; at that altitude, there is not sufficient visibility of the ground to continue the approach visually."

After the accident the radio altimeter was found switched "on" and the range selector set at 400 ft. There was no mention in the Air France operations manual found in the aircraft of the above technical instruction. The operator stated that he was unaware of the technical instruction and, consequently, had taken no action to warn pilots not to rely on the radio altimeter in sandstorm conditions.

Confusion of ASI with Altimeter

The Committee then considered the possibility that the pilot-in-command mistook the ASI for the altimeter. It was noted that his experience on the subject aircraft (Caravelle III) was only about 10 hours compared with approximately 225 hours on the Caravelle VI N. The VI N is equipped with double-pointeraltimeters and single-pointer ASIs, whereas the Caravalle III had single-pointer altimeters and double-pointer ASIs (see figure 24-2). Thus, when the pointers of the altimeter in the VI N are (say) indicating 1 600 ft, they are in the same positions as the pointers of the ASI in a III.

In the past, accidents have occurred as the result of a panel having two different instruments with similar presentation, with the result that the pilot had mistaken one for the other. Pilots have been found to make such mistakes when making instrument approaches in simulators when the workload is high and the conditions under which an instrument approach was carried out were difficult.

In the subject accident, however, the method of presenting information on the airspeed indicator was different from the method of presentation of information on the altimeter. It is to be noted also that the airspeed indicator and the altimeter in both the Caravelle III and Caravelle VI N occupy respectively the same geographical position on the instrument panels, thus making a possibility of error unlikely.

The Committee found it difficult to conceive how such a mistake could have occurred and remained unnoticed by the other crew members, unless they had been reassured by the height being erroneously indicated by the radio altimeter.

The Commnittee also considered whether the NDB radio navigational facility may have provided inaccurate guidance due to night effect or the adverse effect of a static build-up in the sandstorm conditions. However, in view of the use of the facility by other aircraft immediately before and shortly after the accident, the Committee was satisfied that the facility operated efficiently. The Committee was also satisfied that the lack of approach lighting for runway 34 in no way contributed to the accident.

High Velocity Gust

The Committee also considered whether at the time of the accident there was a possibility of vertical or horizontal gusts sufficiently strong to increase considerably che race of descent of the aircraft. It was suggested that this situation could have been aggravated by the aircraft being in its approach configuration, and that it might have been beyond the pilot's capability to control the ensuing rapid descent in the limited height available for full recovery to be effected.

The Senior Meteorological Officer at Bahrain stated that the most severe turbulence in the area of the crash would have occurred between 1700 and 1730 hours GMT and that at the time of the crash he did not believe that any unusually severe turbulence might have existed.

On the other hand, the chief of the Forecasting Centre at Beirut considered that in view of the conditions in the general area, there was a very distinct possibility of horizontal and vertical wind sheer in the Dhahran local area before and after 1930Z in the region from ground level up to 3 000 ft.

At the same time, the department of the Ministry of Aviation, London, was requested to indicate what magnitude and type of gust would be required to cause a Caravelle III air- craft, in approach configuration, to enter an uncontrollable descent. The Ministry of Aviation gave the following opinion:

"According to our calculations, only a gust which led to the aircraft stalling could cause an uncontrollable descent. The aircraft would have been able to recover easily after any smaller gust, and in fact since it was above its minimum drag speed quite a large gust would have been needed even to increase the rate of descent at constant thrust.

The minimum gust which would cause the Caravelle to stall at an airspeed of 154 kt would be an up-gust of about 90 fps EAS. Alternatively, a horizontal gust of about 100 fps could have reduced the airspeed and hence have led to the aircraft stalling. Either of these is, of course, an extremely violent gust."

In view of these opinions, the Committee was unable to determine whether or not a wind shear of sufficient magnitude to cause the aircraft to enter a pronounced rate of descent was present in the area.

2.2 Conclusions


The crew were properly licensed and qualified to conduce the flight and had had adequate rest.

The pilot-in-coormand was more familiar with the manner in which the airspeed and altitude information was presented in the Caravelle VI N, compared with that in the subject aircraft, but the Committee was unable to conclude that this contributed to the accident.

The aircraft had been properly maintained, and was correctly documented, equipped, fuelled and loaded for the flight from Beirut to Dhahran. At the time of the accident the engines were running normally, and there was no mechanical defect or malfunction in the aircraft, its systems or instruments.

Although the reported weather conditions for landing were unfavourable, they did not preclude the pilot-in-command from making an approach to his critical height to determine whether a landing could be carried out in accordance with the operator's weather minima.

It was not possible to determine whether or not there was a local disturbance in the area which produced a high velocity vertical or horizontal gust sufficient to have caused the aircraft to enter a phase in which there was a pronounced increase in the rate of descent.

The air traffic control at Dhahran Airport rendered the aircraft all necessary assistance during its instrument approach, and the NDB and VOR radio navigation aids functioned normally.

The aircraft flew into the sea when descending and slightly banked to the right in an approach configuration It was completing, or had completed, the procedure turn of the NDB instrument approach pattern.

Cause or Probable Cause(s)

The probable cause of this accident cannot be ascertained.

3. Recommendation

That an instrument landing system (ILS) should be installed at Dhahran International Airport.


figure 24-1 approach procedure Dhahran

figure 24-2

airspeed indicator and altimeter

[Back to previous page] [Back to Homepage] [Disclaimer]

Copyright © 1996-98 Harro Ranter/Aviation Safety Web Pages; Updated 12 April 1998