AVIATION SAFETY DIGEST No

j.{ tV\. c ( ( ~ (,.1..( f.\,;5.\ ... I lnc.t' /-\ ,1 ·'"jr. k( COMMONWEALTH OF AUSTRALIA DEPARTMENT OF CIVIL AVIATION

I AVIATION SAFETY DIGEST No. 4

• APRIL, 1955

Prepared by The Accident Published by Authority Investigation and Analysis Division of The Director-General PRINTBD BY HBDG!S 6 BBLL PTY. LTD.. MARYBOROUGH. V ICTORIA ACCIDENT INVESTIGATION SUMMARY

Accident to Vickers Viscount 720, VH-TVA,,at Mangalore Aerodrome, Victoria, on 31st October, 1954 Trans-A ustralia Airlines Viscount 720 VH-TVA crashed after take-off just outside the boundary of Mangalore Aerodrome, Victoria, at 1507 hours on 31st October, 1954. At the time of the accident the aircraft, engaged on routine conversion training, was making a three-engine take-off. The two pilots and one other supernumerary pilot were killed, three other occupants received serious injuries, and the remaining two occupants escaped without injury. The aircraft broke up on impact and was destroyed by fire. The investigation was conducted by the Department of Civil A viation. The following is a summary of the report on this investigation. Australian E astern Standard Time, based on the 24-hour clock, is used throughout this summary. All aircraft speeds quoted are indicated airspeeds, unless otherwise specified.

Circumstances: Two supernumerary pilots and an engineer were standing as observers at the rear of H-TVA, owned and operated by Trans the cockpit, and the remainder of the Australia Airlines, was the first of the occupants were seated in the cabin. VViscount 720 series aircraft received from Vickers-Armstrongs Ltd., Weybridge, After a pre-take-off cockpit check had England. This aircraft arrived in Australia been carried out Nos. 2 and 3 engines were on 13th October, 1954, and was based at opened up to take-off power, the brakes re-. Melbourne. From the date of its arrival until leased, and, as the aircraft moved forward, the day of the accident the aircraft was Captain Fisher gradually advanced No. 1 employed, with the exception of one or two throttle. When the aircraft had travelled scheduled flights, on pilot conversion train some distance a swing to star board develop ing. ed but this was corrected by the use of nose During the morning of the day of the wheel steering. However, this was followed accident VH-TVA was flown for a period of almost immediately by another more severe several hours at Mangalore Aerodrome swing to starboard in which the aircraft left alternate aerodrome for Melbourne Airport the runway and became airborne at a speed -and returned to Melbourne at about 1200 below the minimum control speed. hours. On the afternoon of the same day the aircraf t was again operated at Mangalore As the aircraft left the ground it con on pilot conversion training. Just prior to tinued turning and whilst climbing slightly the take-off in which the accident occurred the starboard wing went down. This turn the aircraft had landed into the north-east steadily steepened and the aircraft, which on three-engines after having completed a had not climbed above 100 feet, steadily lost circuit during which No. 4 engine was de height until the starboard wing tip struck liberately stopped in the take-off, just after the ground. As the aircraft was about to v. speed, to simulate engine failure. On crash one of the supernwnerary pilots, Mc completion of the landing run the aircraft Dougall, moved back into the cabin. was lined up for take-off in the reverse direction with No. 4 engine stopped and the Investigation : pr opeller feathered. Captain Macdonald, the pilot-in-command, was occupying the right Mangalore Aerodr ome, is 450 feet above hand seat and Captain Fisher, pilot-under mean sea level. Runway 22 is a sealed gravel t raining was occupying the left-hand seat. pavement 150 feet wide and 5,880 feet long. The weather on the afternoon of 31st bably full power. No. 4 propeller was fully October, was fine, warm and cloudless, with feathered and there was no indication that unlimited visibility. The wind was light and any attempt had been made to unfeather it. variable. All the evidence, including calculations of the aircraft's performance in such a take Tyre marks left on the runway during the take-off reveal that a swing to starboard off, indicates that the three-engines were de occurred approximately 1,100 feet from the livering full take-off power. _Jl start of the take-off run and, as indicated by No evidence was found to suggest that the marks on the runway, nosewheel steer the flying controls were other than service ing was used to correct it. Marks of a second able and the trim positions were consistent I N OTE. - Terrci.n SLK""roul""\ding, ~ OerOOr""O~ C reJ0+1veJ";,:I .flo+ swing were in evidence 1,800 feet along the with a three-engine take-off configuration. 0 runway and it was during this swing that From the flap selector lever position and the z Ord SOJ'T"\e height OS o.crooror-n the aircraft left the runway at a point some condition of the flaps it is concluded that w J 2,000 feet from the commencement of the flaps were extended 20°. Examination of the er take-off, and became airborne 240 feet pitot system components revealed no defects I- further on. There were no nosewheel stee1· or evidence of malfunctioning. On impact ing marks during this swing. The path of the undercarriage was down and locked but the aircraft is shown in the attached plans. it had been selected "up" approximately 2 seconds prior to impact, which time is in Evidence could not be found of the aircraft sufficient for the selecto1· valve to expose the having struck the ground, or any tree or undercarriage retract mechanism to system object, between the runway and the wreck pressure. The port wheels, tyres and brakes age site. The aircraft was demolished on were found to be in good condition but no impact, the wreckage being spread along a assessment could be made of the starboard distance of 450 feet. Both wings were torn wheel assembly because of the extensive fire off, the fuselage was broken in half and the damage it sustained. The nature of the tyre four engines were separated from the main marks on the runway indicates that there wreckage. The front portion of the fuselage was no malfunctioning of these assemblies, came to rest on its port side with the rear including the brakes, and that the brakes portion, inverted and facing in the oppos.ite direction, lying alongside it. A fire which were not misapplied. occurred on impact spread throughout the The nosewheel assembly sustained re wreckage but, although the area ultimately latively little external damage except for affected was extensive, it did not spread fractures of the lower lugs on the ram-foot rapidly and large portions of the fuselage to which the lower steering link is attached, were not affected by the fire until at least but the cente1·ing helix inside this strut was ten minutes after impact. found broken into a number of fragments. After extensive examination and analysis of LEGEND All the occupants of the cabin survived this damage it was concluded that it had all and reported only moderate deceleration on occurred on impact . · . , impact. One of the two supernumerary • 2 occupants of the flight deck at impact was The nosewheel steering hydraulic circuit • 3 killed, and the other escaped through a hole was found to be intact and serviceable. The G4 in the fuselage, after receiving serious in main hydraulic system sustained extensive juries. Both pilots, who were wearing lap damage and it was not possible to dete1·mine sti~ap type safety belts, but not the shoulder conclusively that this system was operative harnesses available to them, received fatal at the time of impact. head injuries from contact with cockpit fittings. One pilot had no injuries below his Captain D. K. Macdonald, 33 years of age, ACCIDENT TO VISCOUNT AIRCRAFT VH -TVA. head and the other, apart from head injuries, held a first class airline transport pilot had a broken thigh. After an examination licence and his total flying experience amounted to over 11,000 hours, of which AT MANGALORE VICTORIA ON 31 OCT 1954. of all the circumstances it is concluded that 3,158 were in-command of DC-4 type, 1,846 had the pilots been wearing the shoulder CV.240 type, and some 3,000 hours DC-3 LOCALITY PLAN OF CRASH SlTE SHOWlNG harnesses available to them they may have type. He was regarded as a highly com ESTIMATED FLIGHT PATH survived the accident. · petent check and training captain. His ex 5co\e:- \'1 = 1000" An examination of the eng'ines and pro perience on Viscount type totalled 21 hours ------______J pellers established that Nos. 1 and 2 engines 30 minutes at the time of the accident. After were at ·fun power on impact and No. 3 a conversion course, which included ground engine was under substantial power, pro- training in engineering, and flight training

2 3 T. A.S. IN KNOTS & FT/ SEC. J ,, .; g .; v ~ ~ v :.: 1 ~ ~ ~ ~ >Q. i 1 ~ ~ ~ IJ'l -~ " .:!!!.. ii) x ".l :t ~ l< ~ ii:: ~ l( ;:-;.:., x ~' ii:: ~ :t ;( x u: x lfl l< ~ -~ J N <.D ,... Ii) (\j (() ~ II) "' 0

. -0 ·o ·o 0 ·o 0 .8 -~ 0 0 .8 0 0 8 0 0 co

1 ACCIDENT TO VISCOUNT AIRCRAFT VH -TVA. t AT MANGALORE; VICTORIA ON 31 OCT. 1954. RELATION OF TRUE AIR SPEED TO TAKE-OFF RUN. Scole: I" "" 300'

CALCULATIONS BASED ON FOLLOW! NG ASSUMPTIONS. I. THREE ENGINED TAKE- OFF WITH TWO ENGINES TO TAKE-OFF POWER ON BRAKES AND THIRD TO TAKE-OFF POWER WITHIN SIX SECONDS. 2. NO WIND. STANDARD 1. C.A.N. SEA LEVEL CONDITIONS.

4 5 lasting 5 hours 30 minutes, his licence was opening is much greater with the turbo into error by a natural urge to make the likelihood of such an occurrence the success endorsed for the type, on the certification of engine than with the piston-engine. Also, utmost endeavour to avoid damage to this ful use of steering immediately prior to this competence by the Chief Check and Training the response to throttle opening in the turbo aircraft, which was destined to play an last swing cannot be overlooked. Captain of Trans-Australia Airlines. Dur engine is slower than in the piston-engine. important part in his company's operations. ing his conversion Macdonald carried out It would have been obvious to him, once the On the other hand, if the steering several three-engine take-offs. The tyre marks show that the aircraft was aircraft had left the runway, that to mechanism was serviceable it is obvious running 15 feet to the right of the runway from the absence of i·unway marks that the Macdonald was rostered for duty as a centre line, when the fast swing to star abandon the take-off would almost certainly pilots either failed or were unable to use it. training captain on Viscount aircraft, cover board commenced at a point some 1,100 feet have resulted in some damage to VH-TVA, Both were experienced pilots of DC-4 and/ or ing all aspects of this company's Viscount from the start of the take-off. It is by collision with flight strip boundary CV.240 aircraft and accustomed to the use training syllabus after a total experience of estimated that the speed of the aircraft at markers or other features of the aerodrome of nosewheel steering. It is considered that 10 hours 40 minutes on this type. this point would be of the order of 65 knots. outside the prepared flight strip. if either Fisher or Macdonald had been in Marks on the runway indicate that this It is considered that in these circum complete and continuous control each would Captain R. D. H. Fisher, 34 years of age, swing was corrected by nosewheel steering have applied steering instinctively to cor held a first class airline transport pilot stances a training captain fully familiar with and as the swing was controlled these marks the characteristics of the Viscount type rect such a swing. licence and his experience amounted to just became lighter until they disappeared when would not have taken the aircraft into the over 12,000 hours, of which 6, 715 hours had . the aircraft was running straight again. McDougall has said that Fisher corrected been as captain on DC-4 type. Prior to this air. The fact that Macdonald attempted to AltI1ough running straight it was still head do so at a speed below the minimum control the fiTst swing with steering and the absence flight he had completed 4 hours conversion ing slightly to the right away from the run of steering ma1·ks on the runway in the last training on Viscount aircraft in which he speed, indicates that he was not sufficiently swing suggests that Fisher was no longer way centre line. After approximately 200 familiar with this type of aircraft for the had twice taken-off on three-engines. feet in this condition; which would have flying the airccraft, or was flying it in such duties on which he was engaged. It is con a manner as to be deprived of effective taken about 1t seconds, the aircraft again cluded, therefore, that his limited experience Analysis: swung to starboard, left the runway and a steering. At fust it was thought that this materially affected his judgment at this latter possibility might apply if Fisher, who This take-off from which the accident re short distance later became airborne. This time. swing was severe. with his left-hand on the tiller and his right sulted was being attempted in the most Whilst the accident is considered to have on the control wheel, had released the con critical three-engine configuration, i.e. with McDougall, the surviving supernumerary been caused by the decision to take the air trol wheel in order to operate the throttles, No. 4 engine (starboard outer) inoperative. pilot who at this time was standing between craft into the air, this decision was pre thereby reducing the load ori the nosewheels In this configuration, with flaps extended 20° the pilots, states that as the aircraft left the cipitated by loss of directional control dur and rendering the steering ineffective. and the three engines on full power, a mini runway, Macdonald pushed the throttles ing the final stages of the ground run. There mum speed of 96 knots is necessary in order is a number of circumstances which could In tests with another Viscount to explore fully open and pulled the aircraft into the this possibility it was found that,. when the to ensure that, using both rudder and air at an airspeed between 85 and 90 knots have caused the final swing. Some possibi aileron, a constant heading can be maintain definitely not higher than 90 knots. It has lities, such as power plant failure and pro control wheel was released at 75 knots, dur ed. When flying unde1· these conditions at been calculated that with no1·mal accelera peller malfunctioning, have been discarded ing a normal four-engine take-off under the speeds below 96 knots it is not possible to tion in a three-engine take-off the speed of after investigation, but it has not been pos some loading conditions as prevailed in keep the aircraft from turning. sible to eliminate or confirm others. VH-TVA, the nosewheel did not leave the the aircraft where it left the runway would ground, despite a slight up movement of the The take-off safety speed for the con have been 85 to 90 lmots. From what is A puzzling feature of the last swing is nose after which the aircraft still responded ditions existing at the time of this accident known of the characteristics of the aircraft the absence of any nosewheel steering to nosewheel steering. This suggests that if is given in the flight manual for Viscount near the minimum control speed, it is clear marks, which would be expected particularly steering had been applied by Fisher, after VH-TVA as 106 knots and it has been the that it could not be controlled directionally as marks of this sort indicate that nose releasing the control wheel, some steering practice of Trans-Australia Airlines to teach after it left the ground, and its fate was in wheel steering had been used a few seconds marks would have been made on the runway its pilots not to lift the aircraft off the evitable so long as full power was drawn at earlier to correct the previous swing. Hav in the final swing. In any case, Macdonald ground at speeds below 110 knots in a three speeds below 96 knots. In deciding to lift ing regard to McDougall's evidence that he might be expected to push the control wheel engine take-off. VH-TVA into the air at this time, rather was not aware of any difficulty being ex forward again, thereby restoring full steer than abandon the take-off, an error of judge It follows that in a three-engine take-off perienced in the cockpit it is apparent that ing effectiveness, if he saw Fisher release it. whilst the aircraft is still on the ground and ment was committed by Captain Macdonald either the nosewheel steering mechanism It is thought, therefore, that the likelihood at speeds below 96 knots directional control and the accident resulted. failed, unbeknown to the pilots, or that the of Fish.er being deprived of effective steering pilots failed or were unable to use it effec depends on the use of some nosewheel steer Macdonald may have been influenced in is extremely remote. ing to supplement any rudder that may b;i tively, at a stage in the take-off when its this decision by his considerable experience use was essential for directional control. The possibility of Macdonald taking con applied. In order to ensure full effect from on other aircraft recently flown by him, the steering the nosewheels must be held firmly Although the nosewheel and associated trol of the aircraft from Fisher, during or minimum control speeds of which are lower steering mechanism are considered to have immediately prior to the last swing, remains. on the ground by keeping the control wheel than that of the Viscount. It is f elt that in fully forward. been in a serviceable condition at the time In considering this the absence. of any nose an emergency such as faced Macdonald he of the accident, owing to the destruction of wheel tracks in the gravel at the edge of the It is appropriate to mention here that an would be inclined to react automatically some of the hydraulic system plumbing it runway is significant as it indicates that the important characteristic of turbo-propeller under the influence of his predominant ex was not possible to determine this con nosewheels were off the ground before they aircraft, by comparison with piston-engine perience in the DC-4, overlooking the parti clusively. Complete loss of hydraulic pres c1·ossed the edge of the runway. For this to aircraft, is the marked difference in response cular characteristics of the Viscount of sure through such a cause as a broken have happened it is apparent that the con to throttles. The proportionate increase in which he was comparatively inexperienced. hydraulic line could have deprived the pilots trol wheel was previously pulled back and, power obtained in the final stage of throttle It is also possible that he may have been led of nosewheel steering. In assessing the as McDougall has said that Macdonald did,

6 7 in fact, pull it back on taking over control, Trans-Australia Airlines, crashed and this points to Macdonald having assumed burnt approximately one-third of a mile control from Fisher before the aircraft left west of Mangalore Aerod1·ome, Victoria. the runway. This deduction conflicts with The accident occurred when the aircraft McDougall's evidence, but allowance must was flown from the ground following a be made for the possibility of the sequence swing off the runway which occurred of events as recalled by McDougall not being during a practice three-engine take-off. exactly correct in every detail, particularly when the rapidity with which this sequence 2. At the time of the accident Captain R. occurred is appreciated. D. H. Fisher, occupying the left-hand pilot seat, was being given conversion If Macdonald assumed control after the training by Captain D. K. Macdonald. first swing had been substantially corrected, The two pilots and one other super and he could easily have felt this to be numerary pilot were killed, three other necessary because the aircraft was still occupants received injuries and the re heading so as to run off the runway just maining two occupants were uninjured. before it would have become airborne, Fisher would probably take both hands off the con 3. Damage to p1·operty was confined to the trols. McDougall says that when Macdonald aircraft which was destroyed by impact took over he held the control column with and fire. his right hand while he "slapped" the 4. The aircraft was operating under cur throttles forward with his left. It would not rent Certificates of Registration, Air have been possible for him to use the tiller worthiness and Safety and was loaded at the same time and this, together with within the permissible limits. the application of full asymmetric power, ' could have resulted in VH-TVA turning so 5. No evidence was obtained which in- quickly towards nearby obstructions, such dicated any significant defect in the air as elevated runway lights and flight strip craft, its power plants or other equip boundary markers that, influenced by his ment. predominant experience of the handling of aircraft other than the Viscount, he instinc 6. The injuries sustained by the operating crew indicate that they may have sur tively pulled the control wheel back, in an vived this accident if they had been endeavour to fly VH-TVA out of this wearing the full safety harness provided trouble. It has been suggested that with a comparable configuration a DC-4 could for their use in the aircraft. have been taken into the air at 85 knots 7. Training in three-engine take-offs dur without loss of control. It seems probable, ing pilot conversion on four-engine air then, that Macdonald took over from Fisher craft is not a requirement of the De after the first swing and in his doing so partment of Civil Aviation for endorse nosewheel steering was released at a stage ment of such types on pilot licences. in the take-off when its use was essential. Therefore, whilst failure of nosewheel 8. CAUSE: The cause of the accident was steering through loss of hydraulic pressure an error of judgment on the part of the cannot be positively eliminated, its likelihood pilot-in-command in that he took the is considered remote, and it is thought that aircraft into the air at a speed below the the development unchecked of the last swing minimum control speed, following loss arose through nosewheel steering befog re of directional control during the ground leased when Captain Macdonald took over run. control from Fisher; this constitutes a pro 9. Some difficulty in maintaining direc bable contributory cause of the ultimate tional control had been experienced dur accident. ing the take-off run but the final loss probably occurred through nosewheel steering being relinquished as the pilot CONCLUSIONS in-command took over the controls. 1. At 1507 hours on 31st October, 1954, 10. A factor probably contributing to this Viscount VH-TVA, owned by Common accident was the limited experience of wealth of Australia, Australian National the pilot-in-command on this type of Airlines Commission, and operated by aircraft.