Aviation Sri Lanka

Report of the Commission of Inquiry appointed by HE the President to inquire into the causes and circumstances in which Loftleider Icelandic Airways Aircraft DC-*-63F TF-FLA met with an accident in he vicinity of the Katunayake Airport on 15^th November 1978

SECTION I

1. Terms of Reference of the Commission

HIS EXCELLENCY J. R. JAYEWARDENE

PRESIDENT OF SRI LANKA

YOUR EXCELLENCY,

ON November 25, 1979, Your Excellency issued to me a Commission in pursuance of the provisions of Section 2 of the Commissions of Inquiry Act (Chapter 393), with the following terms of reference :

(1) To inquire into and report on the causes and circumstances in which the aircraft bearing registration No. TF-FLA and belonging to Loftleider Icelandic Airways met with an accident in the vicinity of Katunayake Airport at about 23.30 hours local time on November 15, 1978

(2) To consider whether any degree of responsibility for the aforesaid accident may be attributed to any person and

(3) To recommend what steps, if any, should be taken to ensure the avoidance of similar accidents in the future.

2. Synopsis

2.1. On November 15, 1978, Icelandic Airlines Flight LL 001 a DC-8 63 CF (TF-FLA) which was being operated as a charter passenger flight, took-off from Jeddah Airport, Saudi Arabia, at approximately 12.58.03 z to proceed to Surabaya, Indonesia, with a programmed technical stop at Colombo Airport, Katunayake (CAK), Sri Lanka, for fuel and crew change. The aircraft had been chartered by GARUDA Indonesian Airways to carry Indonesian " Haj " pilgrims from Indonesia to Mecca and return. The aircraft contacted Area Control, Ratmalana, at 22.53.24 local time and was informed that the runway in use at Colombo Airport, Katunayake, was 04. The aircraft requested runway 22 and accordingly was cleared for a radar vectored Instrument Landing System (ILS) approach to runway 22. Area -Control who was in contact with the aircraft initially descended the aircraft from FL (flight level) 330 to FL 220 approximately 90 miles out of Colombo Airport, Katunayake. The aircraft was then handed over at 23.06.32 to the Radar Control (CAK) under whose instructions it descended to FL 20 to make an ILS approach to runway 22. The aircraft followed the Controller's instructions and, to all appearances, was making a normal ILS approach to runway 22. The Radar Controller also requested the aircraft to report when it was established on the Localizer, but, though the request was acknowledged, no confirmation was received. The Radar Controller continued to give advisory information on the aircraft's distance and height, the last advisory call being at 23.27.26 when the aircraft was informed thus

" Lima, Lima 001, slightly to the left of centre line, very slightly to the left of centre line, two miles from touch-down, height 650 feet, cleared to land-off this approach."

This transmission was acknowledged by the aircraft at 23.27.37 in the manner " Roger ". There was no further communication from the aircraft.

2.2. Shortly thereafter, the Approach Controller (CAK) sighted the aircraft very low on the approach and called out twice "Lima, Lima 001, you are undershooting ". However, this transmission was not received by the aircraft as the Approach Controller spoke on the approach frequency 119.7 MHz whereas the aircraft was still tuned to the Radar Controller on 119.1 MHz. The Approach Controller observed the aircraft disappearing from sight followed by what appeared to be a ball of fire around the area where it passed out of sight. The aircraft had crashed into a rubber and coconut plantation at a point 1.1589 n.m. from runway 22 threshold, 103.15 feet to the right of the extended centre line of the runway. The aircraft was destroyed by impact and fire.

2.3. The Approach Controller on duty was the first to observe the crash, and the Radar Controller, Area Controller and others were immediately notified of the accident by him.

3. Rescue Activities

3.1. Rescue activities commenced within half an hour of the accident and fire-fighting units were in attendance from this time onwards. Around 5 fire-fighting units were in attendance and the whole operation was coordinated satisfactorily. The main section of the fuselage that was intact was under intense fire and considerable effort was required to bring the fire under control by which time all occupants of this section of the aircraft had succumbed due to fire. Fire-fighting activity was hampered as ready access to the site of the crash was not possible due to the large number of coconut trees that prevented large units from getting closer to the Wreckage.

3.2. The Acting Director of Civil Aviation was personally present with the members of his staff and participated in the rescue operations along with the police and airport staff. The rescue operations were subsequently highly commended by the representatives of the Indonesian and Icelandic Governments.

3.3. All cockpit instruments found among the wreckage were photographed before being handled by anyone and were taken charge of by the Acting Director of Civil Aviation who handed them over to the office of the Defense Ministry to be kept under security. The seals of the packages were broken in my presence at the inquiry.

SECTION II

4. Injuries to Persons

4.1. The injuries to persons were as follows :

Injuries Crew Passengers Others

Fatal 8 175 0

Non-fatal 4 28 0

None 1 46 0

5. Notifications to Interested Parties

5.1. The State of Registry of the aircraft, namely, Iceland, the State of~ Manufacture of the aircraft, namely, the United States of America, the State of maximum number of fatalities, namely, Indonesia, were all informed of the accident. They sent their accredited representatives who made their own preliminary fact-finding investigations and returned to their respective States. The Department of Civil Aviation gave them its full co-operation in the conduct of such investigations.

6. Read-out of Recordings

6.1. At the time of my appointment as Commissioner, only the Icelandic team was still in this country and I had several informal discussions with them in regard to the investigation. Before formal sittings could be held for the recording of evidence of witnesses, it was necessary to send three instruments recovered from the wreckage of the aircraft, namely, the Flight Data Recorder (FD R) (commonly called the Black Box), the Cockpit Voice Recorder (CVR) and the Kifis Box (KB), to appropriate centres abroad for a read-out of the recordings as there are no facilities locally for that purpose. The FDR and CVR were sent to the Air Safety Investigation Branch of the Department of Transport in Melbourne, Australia, and the Kifis Box to the manufacturers in the United States of America.

7. Formal Sittings

7.1. Fifteen formal sittings were held during the period 12th March to 6th April. The records of the proceedings are forwarded separately.

7.2. All the interested parties were given due notice of the formal sittings.

7.3. Mr. Skuli Jon Sigurdarson was present throughout the sittings as the accredited representative of Iceland and participated in the proceedings assisted by Mr. Jon Oltarr Olafsson, Capt. Skuli Br. Steinthorsson an~ Mr. Johannes Jonsson representing the Icelandic Airlines.

Mr. 1. R. Soepartolo was the accredited representative of Indonesia and was assisted by Mr. Soewardi. He was present at most of the sittings and participated in the proceedings.

Mr. D. H. Athulathmudali, Acting Director of Civil Aviation of Sri Lanka was present throughout the sittings and participated in the proceedings.

The State of Manufacture of the aircraft, namely, USA, was not represented at the inquiry though due notice was given.

Mr. V. C. Gunatilaka, Solicitor-General, assisted the Commission as Legal Adviser and Mr. D. J. Rosa, Assistant Director of Civil Aviation (Aeronautical Inspections), as Technical Adviser.

Mr. G. P. S. U. de Silva, Senior Assistant Secretary of the Ministry of Defence, functioned as Secretary to the Commission.

8. Public Representations

8.1. Newspaper advertisements calling for public representations were inserted in the leading local newspapers. A number of letters were received in response to these advertisements but none of them merited consideration. The writers of those letters were not called to give evidence.

SECTION III

9. Navigational Aids

The navigational aids installed at Katunyaka Airport and their status, at the time of the accident are as follows:

9. 1. VISUAL AIDS:

(a) Visual Approach Slope Indicator (VASI).-VASI is a very useful pilot aid. It is of various types. In the type fitted at Katunayake airfield, bars of red and white lights on each side of the runway are so beamed by reflectors that when a pilot is too low he sees all red lights ; when he is too high he sees all white lights ; and when he is on the correct approach path he sees red and white bars one above the other.

The VASI was in satisfactory working order on the night of the accident.

b) Approach Lighting.-The Approach Lighting System gives guidance to aircrafts in the landing phase assisting them in aligning correctly with the runway centre line.

This system had been unserviceable for some months prior to the date of the accident and this fact had been brought to the notice of all airmen by the issuance of a " notam " in accordance with the international practice.

The pilot of flight LL 001 would have been aware of the non-availability of the approach lights. At the time of the accident he was correctly aligned on the centre line by other means and as he was already in sight of the runway lights and of the VASI, the non-availability of the Approach lighting System had no bearing on the accident.

(c) Runway Lights.-The Runway Lighting System was in operation at the time of the accident. The cockpit voice recording indicates that two different voices had mentioned on four occasions that Capt./Co-Pilot " had been visual ". This term implies that ' the Capt./Co-Pilot had seen the runway lights while making the approach to land.

9.2. Radio Navigational Aids:

(a) Very High Frequency Omni-directional Radio Range (VOR).-This was in operation at the time of the accident and there was a remote indicator at the Control Tower to indicate the serviceability of the unit. This aid is of little importance other than to get some rough guidance to align the aircraft with the runway centre line. In the instant case, however, this had been done by the use of the radar.

(b) Distance Measuring Equipment (DME).-At the time of the accident this had been unserviceable for over two months and had been notamed " to that effect. At Katunayake the DME is co-located with the VOR which is not on the extended centre line of the runway. It is useful as a landing aid only if it is co-located with the Glide Path Equipment as is sometimes done in some countries. The function of the DME is for en route navigation purposes. Its unserviceability on the night of November 15 had no bearing on the accident.

(c) Non-Directional Beacon.-The Non-Directional Beacon installed at the airport (NBD-KAT) was in operation at the time of the accident. As the aircraft was correctly aligned on the extended center line of the runway it was of no further importance to the landing of the aircraft.

(d) Non-Directional Beacon at Yakwila (NDB-YKW).-The Non-Directional Beacon at Yakwila is located on the extended centre line of the runway and was in working condition at the time of the accident. The location of this NDB is approximately 17 miles from the end of the runway. As the aircraft was correctly aligned on the extended centre line by other means, it was of little importance to the landing phase of the aircraft.

(e) Instrument Landing System (ILS).-This is by far the most important radio navigational aid associated with the landing phase of an aircraft. It was available to aircraft making an approach to runway 22.

9.3. The ILS comprises the following basic components :

(i) VHF Localizer Equipment (LOC), associated monitor system, remote control and indicator equipment,

(ii) VHF Glide Slope or Glide Path Equipment (GS), associated monitor system, remote control and indicator equipment ;

(iii) Two VHF Marker Beacons, namely, the Outer Marker and Middle Marker (OM and MM), associated monitor systems, remote control and indicator equipment.

The system is operated electronically. The Localizer is a thin beam in the vertical plane and provides correct guidance to align the aircraft on the extended centre line of the runway. The beam comes from a very high frequency (VHF) transmitter at the far end of the runway on the centre line. The pilot sees it as a vertical needle on his ILS instrument.

The Glide Slope (or Glide Path) is a thin beam in the horizontal plane. It provides electronic guidance defining a 3' glide angle and keeps the pilot on the correct descent path. The pilot sees it as a horizontal needle on his ILS instrument.

By flying the aircraft so that the needles are exactly crossed-" locked on the pilot keeps the aircraft on the correct landing approach.

The Marker Beacons located on the extended centre line and away from the airport at a distance of 5 n.m. and 3,500 feet respectively from the threshold ,of the runway 22, provide vertically generated information which can be picked up in the aircraft only when it is overhead of the respective beacons.

9.4. In the Control Tower is situated the remote control and indicator equipment of each of the components which would indicate to the Controller the operational status of the respective components at any time. The indicator would show a green light if the particular component was functioning properly and a red light if it was not. On the night of the accident only the indicator in respect of the Localizer was serviceable. The cable connecting the Glide Slope to the indicator unit was broken and, therefore, the signal that should be received from the monitor of the Glide Slope was not received in the indicator unit. The indicator of the Glide Slope, therefore, constantly showed a red light irrespective of whether the Glide Slope was properly functioning or not. According to the evidence led before me, information as to whether the Glide Slope was functioning properly or not was furnished to the Controller at the tower by a radio technician who was in charge of an ILS portable receiver in a room on the lower floor of the building. It was the duty of that technician to monitor the portable receiver and to communicate immediately to the Tower Controller if the glide slope equipment had shut down or was malfunctioning.

9.5. The cables connecting the two Marker Beacons were also missing (as they were being constantly stolen) and the Controller at the tower had no means of knowing whether they were functioning or not.

On the night of November 15, the aircraft crashed after passing the Outer Marker and before reaching the Middle Marker. Consequently, the status of the Marker Beacons that night had no bearing on the accident.

9.6. One of the matters~ that requires consideration and which became controversial during the course of the inquiry is whether the Glide Slope was working properly on the night of November 15, or whether it was the malfunctioning of the Glide Slope that was the cause or one of the causes of the accident. I shall deal with this matter later on in this report.

SECTION IV

10. Course of the Flight

10.1. The aircraft contacted Area Control Centre, Colombo, at 22.53.24 local time and was informed that the runway in use at Colombo Airport, Katunayake (CAK) was 04 and was also given particulars of the weather. The aircraft inquired whether runway 22 was available. (Runway 22 is the one on which the use of the ILS was available). Area Control confirmed the availability of runway 22 and immediately afterwards at 23.00.48 the pilot confirmed his decision to use runway 22.

10.2. At 23.01.51 meteorological information regarding the cloud base was passed on to the aircraft subsequent to which the aircraft requested the Madras weather. Sometime later, Area Control obtained the Madras weather from Madras and furnished it to the aircraft at 23.18.40.

'10.3. At 23.03.47 the aircraft reported " standing by for. descent " upon which Area Control cleared it for descent to FL 290. At 23.06.09 it was descended further to FL 220.

10.4. Colombo radar took over control of the aircraft around 23.07.00 when it was 90 n.m. out. At 23.10.17 clearance to descend to 7,000 feet was given to the aircraft, and at 23.11.21 a distance call of 60 n.m. was given by the Radar Controller.

10-5. The next radar call was at 23.10.13 when the aircraft was informed by the Radar Controller that it would be a radar vectoring to the ILS on runway 22 and that there was a " bit of weather " on the approach but that visibility was reported to be 6,000 meters. A further clearance to descend to 5,000' was given to the aircraft at 23.17-30 and to 3,000' at 23.22.08. At 23.23.41 it was recleared to 2,000' and a heading of 180 was given.

10.6. In response to an inquiry from the aircraft " Is the ILS working now ? " around 23.24.00 the R/C replied " affirmative " and went on to inform the aircraft " You are closing the localizer from the right, 12 miles from touch-down, recleared to 2,000r.

10.7. Whilst lining up on finals at 23.25.23 the aircraft was informed that it was 8 in. from touch-down and was given a heading change of 220. Seventeen seconds later it was instructed to commence a descent to maintain a 3' glide path with the information that it was 7j miles from touch-down. At 23.25-55 the Radar Controller requested the aircraft to report when it was established on the localizer or when runway was in sight and this call

by the aircraft as " Roger ".

10.8. The next call from Radar was at 23.26.15 after the aircraft had lined up with the centre line of runway 22 when the Radar Controller advised, ,,You will approach the outer marker in 25 seconds". This call was acknowledged by the aircraft as " Roger " at 23.26.28.

10.9. At 23.26.52 the aircraft was informed by Radar that it was 4 miles from touch-down and at a height of 1,300', being cleared to land off the approach to runway 22. This clearance was acknowledged by the aircraft at 23.27.00.

10.10. The next advisory call was given by Radar at 3 miles with height particulars of 1,000' at 23.27.10.

10.11. Radar gave the next call at two miles at 23.27.26 in the following manner, " Lima, Lima 001, slightly to the left of centre line, very slightly to the left of centre line, two miles from touch-down, height 650', cleared to land off this approach." This was acknowledged by the aircraft, at 23.27.37 as "Roger".

10. 12. A further final call to the aircraft by Radar " Slightly to the left of centre line " at 23.27-49 went unacknowledged.

10.13. At 23.28.03 the aircraft crashed 1.1589 n.m. from the threshold of runway 22

11. Impact Sequence and Wreckage

11-1. The impact occurred in an area along the extended centre line of runway 22, the initial contact with coconut-trees being at a height of 163' above mean sea level, 103.15' to the right of the centre line of runway 22. This area was planted with coconut trees, the aircraft brushing the tops of five coconut trees whilst traversing the last 99 feet of the coconut plantation. On leaving the coconut plantation the aircraft entered the rubber plantation almost in a level altitude and passed through the rubber tree tops without an appreciable change in elevation but progressively banking to the Port, the bank angle on leaving the rubber plantation being approximately 20 degrees. Whilst traversing the rubber plantation the aircraft cut a path through the trees approximately 429 feet long and 112 feet wide at the widest point. The Port wing-tip and area immediately after it progressively disintegrated whilst passing through the rubber trees. The aircraft then entered the second coconut plantation and traveled in a slightly descending altitude, the bank to Port increasing progressively up to around 40 degrees over a distance of approximately 396 feet at which point the ground impact marks commenced. The marks on the ground extended to almost 360 feet around which point the aircraft cart-wheeled to the starboard. Whilst cart-wheeling, the Port engines were shed, and the fuselage section from around 12 feet forward of the centre section up to the cockpit sheared away and continued along the path of travel, progressively breaking up into six sections and piling up in one heap approximately 478 feet from the point of initial contact with ground of the aircraft. The remainder of the fuselage, port and starboard wings of the empennage continued to move in a sweeping motion, the tail section approximately 30 feet above ground finally coming to rest almost on the centre line of the runway on a heading 070/290 facing the east. The tail section of the rear galley broke off at this stage and the starboard engines were shed immediately prior to the final resting of the fuselage. A fire ensued in the main fuselage section.

11.2. The port wing-tip and the wing-tip attachment areas were demolished at the initial impact within the rubber plantation. The port wing continued to break down progressively as the aircraft traversed through the rubber and coconut trees up to the point of impact with ground. Other than the port wing, the rest of the aircraft did not suffer any damage up to this point. The breaking up of the fuselage and empennage occurred after the ground impact. During the examination of the wreckage, all flying controls and components were identified ruling out the possibility of any pre-crash failure of the structure. The fire that engulfed the main fuselage section burnt down the fuselage up to window level. There was no fire in the forward area which accounted for most of the survivors being from the forward section.

SECTION V

12. Instruments Recovered and Readings

Although the cockpit area was broken up into sections certain instruments were located and taken charge of, the principal ones being the following :

12.1 Flight Data Recorder (FDR).-The Flight Data Recorder was recovered in an undamaged condition on the morning after the accident from the wreckage strewn around the tail area of the aircraft.

12,2, Cockpit Voice Recorder (CVR).-The Cockpit Voice Recorder was. recovered in a slightly damaged condition on the morning after the accident from the undergrowth around the area where the wreckage of the tail section of the fuselage was scattered. The unit was in a relatively undamaged state.

12.3. Course Indicator (Captain's Panel).-This instrument showed the following readings

(a) Course Indicator-210 degrees

(b) Course Bug set at 220 degrees

(d) Glide Path Needle .5 dots above aircraft position

(e) Glide Slope Flag out of view;

(f) LOC Flag out of view

(g) Compass Flag in view.

12,4. Flight Director Display (Captain's Panel).-The readings as follows :

(a) V-Command Bars-showing marked fly-up ;

(b) Rising Runway-almost in contact with aircraft

12.5. Radio Altimeter (Captain's Panel).-This instrument showed following readings

(a) Flag-out ;

(b) Bug set at 150'

12.6. Pressure Altimeter (Captain's panel).This was set at 1014mbs reading 250'.

12.7. VHFNAV-Captain - 116.3011z

CO-Pilot - 110.30Hz

VHFCOM-Captain - 118-97

CO-Pilot - 131.50

12.8. Flight Director Control Panel

Mode Switch-GA

Altitude Control Switch-off

Pilot Control-0 degrees

13. Read-out of Instruments

13.1. The Flight Data Recorder was taken to the Air Safety Investigation Branch of the Department of Transport, Melbourne, Australia, where a satisfactory read-out was obtained.

13.2. Cockpit Voice Recorder.-This was also sent to the Air Safety Investigation Branch of the Department of Transport, Melbourne, Australia, for a read-out. The cartridge of the Voice Recorder was found in an undamaged condition and was played back on the special equipment available at the ASIB. A copy of the report of the Board is annexed. Of the half hour recording available on the CVR a large percentage of the conversation was in Icelandic. Recordings of the four channels on the cartridge were made individually and collectively and given over to the Icelandic delegation for translation. A certified translation in English of the read-out as furnished by the Icelandic delegation is Annex III. Certain amendments to this certified 'translation were effected by the delegation in March 1979 during the course, of the proceedings.

13.3. Control Tower Tapes.-A recording of the VHF communications between Area Control, Radar Control and the aircraft was available and an accurately timed transcript of this was made out, extracts of which were superimposed on the Final Approach Profile Diagram.

13.4.1. Reconstructed Approach Profile (Annex V).-The Reconstructed Approach Profile was drawn using data computed from the FDR read-outs. The FDR is an old type giving only five parameters and the ground speed cannot be obtained directly. The computation of the Distance Axis (Axis X) on the Approach Profile graph is dependent on accurately knowing the ground speed, which is the vector sum of the indicated air speed (IAS) and the speed of the wind relative to ground (or air speed).

13.4.2. The wind component used was zero as the wind, according to the meteorological report at the time of accident, was " 120 degrees 07 " which meant a 7 kt wind was prevalent from direction 120 degrees. As the approach heading was 220 degrees the component of this wind along the approach path was reckoned to be almost zero.

13.4.3. The nominal glide slope is 3 degrees and the lower broken line on the drawing is the worst assumed glide slope at 1.48 degrees to the horizon. The curve at the top of the drawing is the descent rate. The text appearing above the nominal glide slope in cages is from the CVR transcript and the text below the glide slope in cages is from the Control Tower tapes. The figure in the cages alongside the conversation is local time.

13.4.4. The impact point 'x' is at 163' above mean sea level.

13.5. A cross section of the Approach Profile prepared by the Icelandic delegation with a wind component of 0.

14. Superimposed Transcript

A superimposed transcript was made combining the Control Tower tapes transcription and the CVR transcripts. This appears as Annex VII to this report. The contents of the transcript provided valu4ble, information for the, analysis.

SECTION VI

15. Approach Procedures

The more important of the procedures laid down in the Loftleider Icelandic Operations Manual for Approach and Landing as appearing at pages 4.3.27, 4.3.28, 4.4.18 and 4.4.19 are set out below. The ILS at Katunayake being only of Category 1, the approach procedures set out for automatic approaches under Section P at pages 4.4.19 and 4.4.20 are not applicable.

(a) After passing through 18,000' or transition altitude, select the P.T.C. to " override " and maintain 2,000 setting on the Radio Altimeter until passing 2,000' above the ground and observing the light ON at which time the minimum descent altitude or decision height may be set.

(b) The Co-Pilot will set the Altitude Alerting System to indicate clearance limit altitude throughout the descent and clearance to landing.

(d) Use positive, not excessive rate of descent right down to the minimum descent altitude.

(e) At the outer marker and at 500' above the runway threshold altitude the Pilot not flying the aircraft will cross-check both sets of flights, instruments for proper comparison and ascertain that no warning flags are in view.

(f) If the flying instruments are normal, he will announce 'no flags'. He will also announce the airspeed in relation to Vref (Threshold Speed) and the rate of sink. Example :

Outer Marker 500 ft.

No flags No flags

Ref + 10 Ref +10

Sink 900 Sink 600

(g) Notification will also be made when

(1) passing through 1,000',

(2) leaving 500',

(3) passing through 100' above minima,

(4) at minima, and

(5) approach lights in sight.

Only the altitude need be called out, unless deviation from desired speed, track or glide path is noted.

SECTION VII

16. Failure of the Crew to Adhere to Laid Down Procedures

16.1. A perusal Of the Control communication/CVR transcript and the Approach Profile (Annexes V and V11) indicates that the crew in command failed in many respects to adhere to the procedures laid down.

16.2. The Co-Pilot had not announced the altitude 1,000' prior to reaching the clearance limit altitude (vide 15 (c) above).

16.3. The call-out of altitudes when passing through 1,000', leaving 500', passing through 100' above MDH and at MDH (vide 15 (g) above) had not been made.

The call at MDH is a most important call as this height is the lowest altitude that the pilot descends to if he cannot see the runway and he must stay at this altitude, not lower, until he has visual contact with the runway and, if not visual, he should go around (overshoot) and make another approach.

16.4. At the outer marker and at 500' the standard announcements that to be made regarding

(i) the indication from his scan for warning flags,

(ii) the speed in relation to desired threshold speed, and

(iii) the sink rate

were not called.

The failure to monitor the sink rate was a grave lapse which was a contributory factor to the accident. Considering the average ground speed of the aircraft and its gross weight during descent, the appropriate rate of descent would have been 850/900' per minute. The rate of descent appearing on the top of the Approach Profile (Annex V11) indicates that the rate of descent whilst being on the high side for most of the approach has been well above 1,000'/min. on five peak value excursions, the maximum rate of descent being its high as 2,000'/min. and 1,800'/min. in the final phase of descent.

16.5. The rates of descent of 1,800' to 2,000'/min. are excessive especially at such a late stage on the final approach when the crew had lost visual contact with the runway and were approaching the minima for that runway. This situation could have been avoided if the crew had adhered to the laid down procedures (vide 15 (d)).

16.6. The Icelandic team sought to find an excuse for the failure of the crew to call out the altitudes, the sink rates and Vref deviations by stating that the Co-Pilot was busy complying with the Captain's instructions and had no time to make the aforesaid vital call-outs. An efficient crew member will never miss important calls at critical stages of any approach, however, heavy his work-load may be, as he should know that non-compliance may result in the aircraft and the passengers being placed in jeopardy. it is not clear why the Company procedure does not provide for the Flight Engineer being utilized to monitor important procedures and call-outs during approach when the Co-Pilot is busy otherwise. In any event, if the procedures had been strictly followed, there should have been no clash between the Co-Pilot's carrying out the Captain's orders and making the call-outs expected of him.

SECTION VIII

17. Information Furnished by the Radar Controller

17.1. The point was raised that there was discrepancy in the position data passed on to the aircraft by the Radar Controller ; that when the R/C gave the call " 4 miles out at 1,300' " the position of the aircraft according to the Approach Profile (Annex V11) was 4.2 miles out at 1,640' ; similarly, when the call was " 3 miles out at 1,000"' the position of the aircraft was 33 miles out at 1,290' ; and when the call was " 2 miles from touch-down at 650"' the position of the aircraft was 2.77 miles out at 1,020' ; and that the Captain's excessive rates of descent at those points were probably due to his anxiety to conform to the calls given by the R/C. It was thus sought to lay the blame on the R/C for the excessive sink rates at those points. It should be noted that the excessive sink rates were not confined to those points alone. Apart from that, one may consider whether the blame for the excessive sink rates can be reasonably passed on to the R/C.

17.2. The aircraft was cleared for an ILS approach to runway 22 by the R/C at 23.16.13 and the R/C informed the aircraft that it was a " radar vectoring to ILS ". The principle of a radar vector to the ILS is " to provide radar vectoring of arriving traffic on to pilot-interpreted final approach aids " (ICAO DOC 4444 Rules of Air Traffic Services, P. 10). The radar vector to ILS positively terminates once an aircraft is established on the ILS. The Captain, therefore, once he was established on the ILS should and would have known that further radar vectoring was unnecessary and superfluous ; that he was no longer under the control of the R/C and that he was not obliged to take note of any advisory information given to him by the R/C.

17.3. The R/C had at 23.25.25 instructed the aircraft to report when it was established on the localizer or when runway was in sight. The crew had however, failed to report at any stage that they were established on the localizer or that runway was in sight and consequently the R/C appears to have continued ,to give advisory heights. This subsequent advisory information was definitely not a part of the radar vectoring to the ILS as the aircraft was already established on the localizer. Had the aircraft reported established on the localizer or had the runway in sight the R/C would have terminated the vectoring and handed -over the, aircraft to the Tower Approach Controller for the final approach and touch-down. It was primarily the Captain's failure to report that he was established on the localizer that was responsible for the aircraft not being handed over to the Tower Approach Controller at the proper time. Apart from the fact that he was not under any obligation to take note of the superfluous advisory information that was continuing to be furnished to him by the R/C, the Captain could have checked his own altimeters before accepting the heights furnished by the R/C and acting on them. The Pilot should have known that it was not a Surveillance Radar Approach (SRA) that he was following. If the readings on his altimeter did not tally with the information furnished by the R/t, the R/C's information should have been ignored. It was also open to the Captain to bring to the notice of the R/C that the altitudes furnished by him did not tally with the readings on his altimeters and to have asked for confirmation. In all the circumstances, I do not think it reasonable pass on to the R/C the blame for the excessive sink rates of the aircraft.

17.4. The Icelandic team marked in evidence a Flight Path cross section with wind component + 10) and submitted that the aircraft was always farther away from the runway touch-down point than specified by the Radar Controller. According to them, when the Radar reported the aircraft to be 4 n.m. from touch-down and the altitude to be 1,300' the aircraft was actually at 4.5 n.m. and at an altitude of 1,530' ; when Radar reported 3 n.m. and an altitude of 1,000' the aircraft was about 3.7 n.m. at 1,200'; and when Radar reported 2 n.m. and the altitude to be 650' it was in fact at 2.8 n.m. and 870' altitude. It will be noted that the figures given by the Icelandic team on the basis of Annex VIII differ from the figures on the basis of the Approach Profile (Annex VII), perhaps due to the difference in the wind component. There is no reliable data in regard to the wind component at the relevant times and the accuracy of the distances on the approach profiles that are reconstructed cannot therefore be completely depended upon.

17.5. The Icelandic team, however, submitted that the erroneous distance and altitude information provided by the Radar Controller was a significant contributing factor to the accident.

17.6. On his last call the Radar Controller had indicated that the aircraft should be at a height of 650' and 2 n.m. from the runway. Even assuming that the heights and distances furnished were not accurate, the pilot's descent below the last call was on his own responsibility. If he had descended from the altitude at which he was at a normal sink rate and at the decision height of 250' (or 228) had initiated an overshoot if the runway was not visible, the accident would not have taken place. Any wrong advisory information given earlier by the Radar Controller could not, therefore, have been a contributory cause of the accident.

SECTION IX

18. Radio Altimeter

18.1. The obstruction clearance limit (OCL) for an ILS approach to runway 22 at Katunayaka is 200'(vide Annex IX). Consistent with this height was the instruction of the Captain to set the Radio Altimeter bug at 250' as seen at 23.26.45 on the transcript (Annex VII). Whilst no call-outs had been made approaching the MDH, the Radio Altimeter (RA) found on the Captain's panel of instruments recovered from the wreckage showed the bug set at 150' -vide the photograph, Annex X. It is not possible to determine at what stage the bug on the RA had been set to 150'.

18.2. The Icelandic team submitted that since the Captain had at 23.26.45 instructed the Co-Pilot to set the Radio Altimeter at 250', he would have set his own too at 250' ; that the knob with which the bug has to be set is very easily moved that turning the knob half a turn will change the setting by one hundred feet and that, normally only a slight touch of the knob is enough to turn it. For these reasons they were of the opinion that the bug had moved during or after the crash.

18-3. In view of the above submissions, I have carefully re-examined the Radio Altimeter and tested the knob and the bug. The knob is undamaged and is turned by a rotary movement. In order to change the setting of the bug by one hundred feet, the knob has to be turned one full turn and not half a turn. A half turn changes the setting by only fifty feet. The bug cannot be moved except by a deliberate manipulation of the knob, unlike the other instrumentation on the panel which have spring loaded indicator needles where the tendency is for the needles to return to the zero position on power cut-off or the possibility exists that they may be shaken round due to forces of impact, thus settling in a completely different position from that indicated while it was functioning properly. It is not correct that a slight touch is sufficient to turn the knob. In my opinion, it is highly improbable that a full turn of the knob to change the setting from 250' to 150' could have taken place as a result of the impact during the crash. It seems to be much more likely that the Captain had, by error, set the bug at 150' instead of at 250', though he intended to set it at 250'.

18.4. On the other hand, if the Captain had correctly set his radio altimeter bug at 250', the warning light would have come on when the aircraft came down to that height. If he was scanning his instruments, he could not have failed to notice that fact. In that event, one cannot understand why he did not overshoot if the runway was not within view.

18.5. None of the instruments on the Co-Pilot's panel were recovered as they were all badly smashed up. It is possible that the Co-Pilot's radio altimeter had been set at 250' but, perhaps, the Co-Pilot was too pre-occupied with looking out, watching for the runway lights, that he failed to take note of the warning light on his panel when the plane descended to 250'.

18.6. An erroneous misreading of the altimeter by the crew is not unknown. For example, in the aircraft crash that took place at Escambia Bay in Florida on May 8, 1978, the Captain and the First Officer both admitted at the hearing that they had misread the altimeter reading. In that case too there had been no altitude call-outs. The report of the said air accident by the National Transportation Safety Board (NTSB) dated November 9, 1978, states at page 19:

" The Captain and First Officer testified that they misread their barometric altimeters during the latter stages of the descent after they were Cleared to descend from 1,700' . . . . The Captain said that he misread his altimeter at 500' and believed he saw 1,500' . . . . . The First Officer Wd that he failed to make the required altitude call-outs because he was never aware of the fact that the aircraft was 1,000' until just before the Impact."

J. N. Ramsden in his book " The Safe Airline " (MacDonald and Jane's London, 1976, page 207) says :

" Altitude awareness is perhaps the professional pilot's most highly developed facility instilled into him from the first hour of training. But in the first half of the 1970s there were more than 80 fatal approach accidents to public transport aircraft, with the loss of over 2,600 lives. Most of these accidents were caused by the crew's unawareness, until too late, of their proximity to the ground."

18.7. According to the evidence, the decision height at Katunayake in terms of the procedure laid down by the Icelandic Airlines is 228'. The Captain appears to have been cautious and decided to fix it at 250' for the landing, taking into account, perhaps, the stormy weather. (It was stated in evidence that Air Ceylon pilots usually fix 300' as the Decision Height especially in bad weather). If the altimeter bug had been erroneously set at 150' the warning lights would not have come on at 250' to warn the Captain that he was at the Decision Height. In the absence of altitude call-outs and of the warning lights the Captain was probably not aware of the altitude when he allowed the aircraft to go down below the Decision Height and to reach a dangerous level so as to hit a tree which was 163' above mean sea level. This would also confirm that there had been no proper cross-check of the flight instruments by the crew.

18.8. It was also stated that the Icelandic crew during Category One approaches utilize the Radio Altimeter only for guidance and cross-check of the Barometric Altimeter and it is the Barometric Altimeter that is used by the pilot to establish his Decision Height. The Barometric Altimeter, however, does not have a warning light on the Captain's panel and it is only the Radio Altimeter that would have given him the warning in the absence of call-outs by the Co-pilot.

18.9. Since all altitudes of the Ground Proximity Warning System (GPWS) mode IV are computed from the aircraft radio altimeter (Vide Ranisden : Ibid, p. 210), the GPWS too would not have given any warning until the aircraft came down to the altitude of 150'.

18.10. It seems clear that both the Captain and the Co-Pilot became aware of the dangerously low altitude to which the aircraft had descended only when the Co-Pilot saw and announced that the VASI lights were red.

SECTION X

19. Decision Height

19.1. According to ICAO, Decision Height is the height below which an aircraft on an electronic glide slope may not descend, and at which an overshoot must be initiated if there is no visual reference. (Vide Ramsden : Ibid, p. 208) In the instant case the Pilot at the height of 250' (which he appears to have fixed as the Decision Height) or at least at a height of 228' (which, according to the Icelandic Airlines' laid down procedure was the Decision Height for Katunayake) should have initiated an overshoot if the runway was not visual at that stage. Had he done so, whether he was flying a glide slope or not, he would have avoided a crash. The fact that he was flying the glide slope was no justification at all for him to descend the aircraft to a level below 2j8'. It was stated by the Icelandic team that according to the laid down procedure of the Icelandic Airlines the Pilot was strongly recommended " to remain on instruments " until he reached the altitude of 50' over the-threshold of the runway. It was submitted that the Captain therefore acted in conformity with instructions in flying the glide slope even below the Decision Height level. An examination of the provisions of the Icelandic Airlines' Operations Manual for DC-8-63 aircraft at p. 4.4.19 shows that the recommendation referred to is applicable only to automatic approaches. Category I ILS is not meant for automatic approaches and a pilot should not rely on the ILS below Decision Height.

19.2. In any event, the expression " remaining on instruments " would mean a scan of all the instruments on the Captain's panel in the cockpit and not merely 'flying the ILS'. Had the Captain had a proper scan of all instruments he could not have failed to detect the low altitude to which the aircraft was descending. The ILS at Katunayake falls under Category I and is not meant for an ILS descent below the Decision Height. The Glide Slope cannot be used as a touch-down guidance aid. (vide Avionics Navigation Systems by Myson Kayton and Walter R. Fried, p. 532).

19.3. Captain S. R. Wickramanayake, a Pilot of considerable experience in flying all types of aircraft and who is at present Chairman of Air Lanka, stated in evidence that the ILS category I is not designed to bring the aircraft

down to the threshold and that at the Decision Height, if the runway was not visual, the Pilot had to initiate an overshoot. To a question put by Captain Steinthorsson of the Icelandic team, " Do you agree with me that if you Initiated a missed approach at Decision Height, you can very well slip about a few feet ? " he gave the answer, " Forty to fifty feet is allowed ". So that, if the Pilot had initiated missed approach procedure at the height of 250' (which he appears to have set for himself as the break-off point) he should still have been able to avoid the crash which took place at an altitude of 163'.

SECTION XI

20. Was a Bent Glide Slope the Cause of the Accident ?

20.1. The Icelandic delegation produced in evidence a flight path cross section (Annex VIII) prepared by them and pointed out that according to that cross section the aircraft had followed the glide slope at the time of the accident. They submitted that the Captain's course indicator found in the wreck indicated that at the time of the crash the aircraft was receiving ILS signals and was slightly to the right of the localizer which coincided with the spot where the crash took place and only slightly low on the glide path (approximately J "dot"). According to them, the glide path was bending downwards approximately 3-5 n.m. from the touch-down zone and it was by following that bent glide slope that the Pilot came down to a dangerously low altitude and crashed.

20-2. In support of their submission they relied on the following:

(a) The flight path cross section referred to above

(b) Memo of a meeting with a Mr. Heyn (AC 1)

(d) Certain entries in the ATC Log Book (X 13) and the extracts from the same (AC 11) ;

(e) The Ground Proximity Warning System did not alert the Pilots that the aircraft was below the glide path and that from the Pilot's point of view the approach continued to be normal until he was alerted by the Co-Pilot's call that the VASI lights were red.

20.3. As regards the memo of a meeting with Mr. Heyn it was stated that Mr.Heyn is attached to the Flight Inspection Branch of the FAA and that he had expressed certain views in regard to the formation of bends in a glide slope when a team from Iceland met him in the U.S.A. for consultations. Mr. Heyn was, however, not called as a witness before me and I informed the Icelandic delegation in the course of the proceedings that opinions expressed to them by any person, however eminent he may be, would not be relevant evidence and cannot be acted on by me unless that person was called to testify personally at the proceedings before me or his opinions were supported by any competent witness who gave evidence before me.

The document marked AC 7 is also one based on the opinion expressed by Mr. Heyn and cannot be availed of as relevant evidence.

20.4. Mr. Heyn appears to have expressed the opinion that significant deviations can occur in ILS glide slope and localizer beams as a result of " improper maintenance procedures " and that those deviations can be aggravated during inclement weather, such as heavy rainfall. On the evidence placed before me it is not possible to come to the conclusion that the ILS at Katunayake had been improperly maintained.

20-5. But the Icelandic team relied on certain entries in the Log Book of the Tower Controllers to show that the ILS had not been working satisfactorily. The principal officer in charge of maintaining airport equipment and all navigational aids at Katunayake is Mr. Somasiri who has been attached to the Department of Civil Aviation for 20 years. He had had training at the Air Services Training School in Canada on radar fundamentals, VOR equipment and test equipment. He had also had practical training at Halifax International Airport. In 1972-73 he had attended the Federal Aviation Administration Academy in Oklahoma City, USA, and had had training on navigational aids for 7J months. He had also had a period of training in Manila where an ILS ide