Aviation Accident Report: United Air Lines Flight 823
CIVIL AERONAUTICS BOARD
AIRCRAFT ACCIDENT REPORT
UNITED AIR LINES, INC.
VICKERS VISCOUNT 745D, N7405
NEAR PARROTSVILLE, TENNESSEE
JULY 9, 1964
A United Air Lines, Inc., Vickers Vicsount 745D, N7405, Flight 823, crashed 2-1/4 miles northeast of Parrotsville, Tennessee, at 1815 e.s.t., July 9, 1964. Thirty-four passengers and the four crewmembers died in the crash. One passenger died of injuries following a free fall from the aircraft before the crash. The aircraft was destroyed by fire and impact damage.
Flight 823 was a regularly scheduled operation from Philadelphia, Pennsylvania, to Huntsville, Alabama, with en route stops at Washington, D. C., and Knoxville, Tennessee. The flight operated without any reported discrepancies or difficulties until approximately 1810 e.s.t., when it was observed flying at low altitude trailing smoke. The flight continued in a southwesterly direction and at a point approximately 1.6 nautical miles before the impact site, a passenger was seen falling from the aircraft, and a short time later a cabin window was seen falling. The aircraft was then observed going into a nose-high attitude, the left wing and the nose went down, and the aircraft dived into the ground, exploded, and burned.
The Board determines that the probable cause of this accident was an uncontrollable inflight fire of undetermined origin, in the fuselage, which resulted in a loss of control of the aircraft.
The aircraft, N4705, operated as United Air Lines (UAL) Flight 609 from Raleigh-Durham, North Carolina, to Philadelphia with intermediate stops at Washington, D. C., Buffalo and Elmira, New York, and Williamsport, Pennsylvania. The aircraft departed Raleigh-Durham at 0705 and arrived at Philadelphia at 1245. The captain and senior stewardess of the flight from Buffalo to Philadelphia stated that the flight was routine and the aircraft operated normally.
The captain, first officer, and one stewardess served aboard UAL Flight 662 from Washington, D. C., to Philadelphia, Pennsylvania, arriving at Philadelphia at approximately 1215.
United Air Lines Flight 823, a Vickers Viscount 745D, N7405, departed Philadelphia at 1513 arriving at Washington, D. C., at 1554. No discrepancies were reported by the crew at Washington and no maintenance, other than servicing, was required or performed prior to departure for Knoxville.
Flight 823 was a regularly scheduled domestic public transportation flight from Philadelphia, Pennsylvania, to Huntsville, Alabama, with intermediate stops at Washington, D. C., and Knoxville, Tennessee. Flight 823 departed Washington, D. C., at 1636 with an estimated arrival time of 1813 at Knoxville, Tennessee.
An Instrument Flight Rule (IFR) flight plan, filed for Flight 823 from Washington to Knoxville, requested routing via Victor Airways 143, 140, and 16 at an altitude of 14,000 feet m.s.l. Following departure from Washington with 35 passengers and a crew of four, the flight operated without any reported difficulties to the Holston Mountain VOR. The crew reported to the Atlanta Air Route Traffic Control Center (ARTCC) over that fix at 1758:35 and estimated their arrival at Knoxville at 1821.
Approximately one minute after having reported passing Holston Mountain, the crew requested a clearance to descend to the lowest available altitude. They were cleared to descend to and maintain 8,000 feet. Three minutes later the crew canceled their IFR clearance. The controller offered to pass control of the flight to Knoxville Approach Control when they were closer in and advised they could stay on the Center frequency. At 1802:55 the crew responded to this transmission with "OK." This was the last known transmission from the aircraft.
The aircraft's radar target disappeared from the controller's scope at 1813:30. At 1814, after waiting four sweeps of the radar antenna, the Atlanta controller called the aircraft to advise them he had lost radar contact but he received no reply.
Numerous ground witnesses observed the aircraft flying at what they considered to be a very low altitude. Witnesses estimated the altitude of the aircraft to be from 200–500 feet above the ground along a line generally parallel to, but south of, V16. The last 10 to 12 miles of the flightpath were observed by a number of ground witnesses, several of whom stated they saw smoke of varying density apparently coming from the fuselage of the aircraft.
Flight 823 was first observed approximately 36 miles southwest of Holston Mountain VOR at an estimated altitude of 5,000 feet descending. There was no visible difficulty at that time. A witness who observed the aircraft from a position 11 miles northeast of the crash site was the first to report seeing anything unusual. She noted a violet red light burning on the fuselage. She could offer no further clarification as to the location of this light. While she could read the company name on the side of the aircraft, she did not see any smoke. The time was about 1810 and the aircraft was estimated to be at an altitude of 500 feet.
The first witness to report smoke from the aircraft was approximately five miles from the crash scene. He stated that "...smoke was coming out of the tail part..." and "...there were brown spots like the paint was off of it about half way back on the body...." Witnesses about one mile farther along the flightpath did not observe anything unusual, except the low altitude, until the aircraft had passed them at which time they observed smoke coming from the aircraft.
A number of witnesses about two miles from the crash site, near the flightpath, did observe black smoke coming from aircraft fuselage near the wings. A large black object, later identified as a passenger, was observed to fall away from the aircraft, followed by dense black smoke. The witnesses stated that the object did not strike the tail of the aircraft after coming out the left side over the wing. Farther along the flightpath a bright object, later identified as the left No. 9 emergency cabin window, fell from the aircraft. Heavy smoke was seen continuing to come from the aircraft.
While the majority of the witnesses did not report seeing any fire, some witnesses did report signs of fire in or on the aircraft.
Shortly after the passenger and the window fell from the plane the aircraft nosed up, the left wing went down, the aircraft nosed down, and crashed into a rocky wooded hillside.
The crash occurred approximately 41 nautical miles east-northeast of the Knoxville VORTAC and about 2-1/4 nautical miles northeast of Parrottsville, Tennessee, at approximately 1815. The accident occurred during daylight hours at an elevation of approximately 1,400 feet m.s.l.
The four crewmembers and 34 passengers were killed in the crash. One other passenger died of impact injuries after a free fall from the aircraft, landing about 1.6 nautical miles from the crash site.
The aircraft was destroyed by impact and fire damage.
Private property damage was confined to burned trees and ground cover.
Captain Oliver E. Sabatke, 41, held airline transport pilot certificate No. 479853 with ratings in DC-3, DC-3S, DC-6/7, Lockheed Constellation, and Vickers Viscount. His last Viscount proficiency training was completed June 30, 1964, in the Viscount simulator. His last physical examination was completed April 4, 1964, with no limitations. His original checkout in Viscount aircraft was May 18, 1957. Captain Sabatke had a 24-hour off duty period prior to his first flight on July 9, 1964, and had been on duty 6:19 hours at the time of the accident.
First Officer Charles L. Young, 37, held airline transport pilot certificate No. 721324 with a Vickers Viscount rating. He passed a first officer's proficiency check in the Viscount February 7, 1964, and his last en route check July 7, 1964. His last physical examination was completed January 14, 1964, with a limitation attached that "Holder shall wear correcting glasses while exercising the privileges of his airman certificate." He had a total of 7,715 hours flying time including 2,100 hours in Viscounts, and had passed a type rating check in the Viscount February 14, 1963. He had a 25-hour rest period before reporting for duty July 9, 1964, and had been on duty 6:19 hours at the time of the accident.
Stewardess Carole L. Berndt, 21, was employed by United Air Lines, Inc., February 14, 1963. She had completed her last recurrent training May 12, 1964, and had 14:08 hours rest period before the flight.
Stewardess Virginia K. Vollmer, 22, was employed by United Air Lines, Inc., March 11, 1964, and completed her last competency check April 20, 1964. She had an 18:04-hour rest period prior to this flight.
N7405, a Vickers Viscount 745D manufactured November 14, 1955, with manufacturer's serial No. 103, was owned by United Air Lines, Inc. The airframe had flown 23,804 hours and had undergone a numbered inspection (#2) 180 hours before the accident. The aircraft received a preflight check six hours before the accident and was given an en route preflight inspection just prior to takeoff from Washington.
The aircraft was powered by four Rolls Royce model Dart 510 engines equipped with ROTOL Ltd. propellers model R130/4-20-4/12E with RA 25842 blades installed. Total times of engines and propeller hubs were:
All the propeller blades had the same TSO as their hubs. The TT on the blades of Nos. 3 and 4 propellers were identical with their hubs but the total time of the blades on No. 1 propeller was 15,312:58 and on No. 2 the propeller was 20,376:28.
A review of the maintenance records revealed there were no known discrepancies when the aircraft left Washington. The records indicated that N7405 was maintained in accordance with existing United Air Lines, Inc., and FAA approved directives with the exception of the deletion of a turbine blade inspection of No. 2 engine during the last block overhaul, and a similar deletion of an inter-cooler air inlet cleaning operation. The aircraft and powerplants were reported to be airworthy at the time of departure from Washington.
The takeoff weight of Flight 823 was 58,948 pounds and the weight was estimated to be 51,468 pounds at the time of the crash. Maximum allowable takeoff weight for runway 33 at Washington National Airport was 60,600 pounds. The center of gravity (c.g.) limits for the flight were 9 percent Standard Mean Chord (SMC) forward and 26 percent SMC aft. The aircraft was within these limits at takeoff and was computed to be within limits at the time of the crash.
The aircraft was loaded with 10,900 pounds of Jet A at takeoff and was computed to have 6,327 pounds of fuel aboard at the time of the crash.
The United States Weather Bureau (USWB), Washington Office, issued an aviation area forecast at 1345 valid for 12 hours beginning at 1400, which included Washington D. C., eastern North Carolina, and Virginia. A trough was reported extending southwestward from a low off the New Jersey coast. Throughout the area clouds were forecast at 3,000 feet scattered variable to broken; ceilings 10,000 feet broken variable to overcast; scattered ceilings of 1,500 feet broken, 5,000 overcast; visibility two miles in light rain, showers, or thunderstorms with light rain showers; becoming 3,000 feet scattered, ceilings of 12,000 broken and eight miles visibility by 2200. Thunderstorms were forecast to be more numerous in West Virginia through western Virginia and western North Carolina. The freezing level was forecast to be 11,000 to 12,000 feet with brief moderate, possibly heavy icing above the freezing level in heavy showers and thunderstorms. Brief severe turbulence was forecast to be likely in thunderstorms.
The USWB office at Memphis forecast squall line thunderstorms over western Tennessee from 1500 to 2100. Central and eastern Tennessee were forecast to have ceilings at 2,000 to 3,000 feet broken, variable to scattered, 10,000 overcast variable to broken with layers merging and tops to 40,000 feet and, in scattered thunderstorms and heavy rain showers, ceiling 500 feet overcast, visibility one mile. Heavy mixed icing was forecast in thunderstorms above the freezing level at 12,000 to 13,500 feet.
The Knoxville terminal forecast issued by USWB Memphis at 1145 valid for 12 hours beginning at 1200 called for scattered clouds at 3,000 and 30,000 feet with a chance of a thundershower.
The crew was provided with the latest available weather sequence reports for their route before departure from Washington.
The ground witnesses reported that the weather in the incident area, and along the last segment of flight from Holston Mountain, was clear and calm, with a few high scattered clouds. The accident occurred in day VFR conditions about 1:40 before sunset.
There were no reported discrepancies of ground or airborn navigation equipment during the flight. The aircraft came under radar observation and control the Atlanta Air Route Traffic Control Center at 1735. Both primary and secondary radar targets were observed until they disappeared at 1813:30, some 15 minutes after passing Holston Mountain VOR. The center's radar had been flight checked as usable down to 6,000 feet m.s.l. along Victor 16 from Holston Mountain VOR to the Ottway Intersection 39 miles southwest of the VOR. Radar beacon targets had been observed along this portion of Victor 16 down to an altitude of 4,000 feet m.s.l. The last altimeter setting given the crew of Flight 823 was 29.87 inches for Knoxville by Atlanta Center.
The Knoxville VORTAC was operational during the period Flight 823 was flying between Holston Mountain and Knoxville.
There were no reported discrepancies in air to ground communications during the flight from Washington to Holston Mountain.
The flight was observed on radar to pass Holston Mountain VOR at 1757. The crew called at 1758:35 and reported passing Holston Mountain 1758. At 1759:45 the flight requested the "lowest altitude available" and was cleared to descend to and maintain 8,000 feet. At 1802:45 the flight cancelled their IFR clearance. At that time the Atlanta controller offered to hand the flight off to Knoxville Approach Control when it got a little closer in, if they would stand by on the Center frequency. At 1802:55 the flight acknowledged this transmission with "OK." This was the last known transmission from the flight.
At 1814:10 the Atlanta controller called the flight to report his loss of radar contact but received no reply. All other attempts by him to contact Flight 823 were fruitless.
There was no evidence of an emergency or any unusual situation in any transmission by the crew.
Not involved in this accident.
The aircraft was equipped with a Lockheed Aircraft Service model 109C flight recorder, serial No. 578. The recorder was installed in the electrical compartment below the cabin floor at fuselage station (FS) 389. The recorder shell was crushed to one-half its original diameter and had separated due to shear loads. The contents were exposed to fire which destroyed the aluminum recording tape. No usable information was available. This is the first instance of the destruction, by fire, of recording tape in a Lockheed recorder.
The aircraft struck on a 45-degree, heavily wooded slope at an elevation of approximately 1,400 feet m.s.l. The heading on impact was 135 degrees, the nose was approximately 55 degrees below the horizon, and the bank angle was about 45 degrees left wing down. The wreckage, except the No. 9 emergency window and some small pieces from the cabin interior, was contained in an area 300 feet long and 200 feet wide.
A 1/2 to 3/4 mile wide ground search was conducted for 5-1/2 miles back along the flightpath from the impact site. The No. 9 emergency window, scraps of cloth from the cabin interior, an emergency exit placard, and parts of a window seal were located 2,320 feet, on a magnetic bearing of 035 degrees, from the primary wreckage site. The cabin material was scattered over an area which extended about 600 feet from the window. The free-fall victim was located 8,400 feet, on a magnetic bearing of 030 degrees, from the primary wreckage site. A cigarette lighter with a clear plastic fuel reservoir 7/8 full was found near the body. No other material from the aircraft was found along the flight path.
The aircraft wreckage was fragmented and severe ground fires burned for several hours after the accident. All major components were accounted for at the accident scene and there was no evidence of pre-impact structural failure. The landing gear and flaps were retracted. The primary flight control trim settings were: aileron, one degree right wing down; elevator, one degree nose down; and rudder, 3/4 degree, nose right.
The engines were rotating with the propeller blade angles in the flight range at the time of impact. The powerplants revealed no evidence to suggest failure or maloperation.
The hydraulic and electrical systems components, including lines and wiring respectively, were extensively damaged by impact and, in part, destroyed by fire. However, an examination of the pieces that were recovered did not reveal any evidence of system failure or malfunction. One electrically operated fuel boost pump and the NESA inverter showed rotational damage.
The flight control system utilizes push-pull rods beneath the cabin floor, located to the left of the fuselage centerline. Except for terminal ends and some steel sections, these rods are made of aluminum. Parts of the aluminum material were destroyed by fire.
Despite impact and fire damage, some information was obtained from the navigation and communications equipment. The two VHF radio transmitter selector switches were positioned on "No. 1 VHF Com." The No. 1 VOR navigation receiver was tuned to the Knoxville VORTAC. The No. 2 VOR receiver could not be identified. The omni bearing indicators read 066 and 148 degrees. The cockpit control head of the Air Traffic Control (ATC) radar beacon transponder was recovered positioned on Channel 11, Mode B; however, the Atlanta ARTCC radar had been receiving Channel 11, Mode A. Of the two automatic direction finding loop antennas, only one provided bearing information, 096 degrees. The distance measuring equipment (DME) as recovered was tuned to the Knoxville VORTAC with the mileage module locked at 47.5 miles. One altimeter was recovered, set at 29.82.
Both inflight and post impact fire occurred in this accident. The extent of the post impact fire as well as the extensive break-up hampered the investigation with respect to origin and progress of the fire. However, a comprehensive mock-up did permit some determinations in these regards.
Ground witnesses established by observations of smoke that there was an inflight fire. Burns and soot deposits on the free-fall victim and fire damage to bits of cabin material that fell away from the post impact fire area located inflight fire in the passenger cabin of the aircraft.
Examination of the wings, empennage, and powerplants did not reveal any evidence of inflight fire.
The fuselage beneath the cabin floor and rearward of the main spar (FS-414) contains the fuel burning Janitrol heater and is also the section through which the engine fuel system cross-feed line passes. The fuel line runs within a shroud which is vented to the outside. In the event of a double leak involving both the fuel line and shroud, fuel would not be released into the fuselage during pressurized flight. The Janitrol heater is not used in flight. There was no evidence of inflight fire in this area. This is the only portion of the fuselage that has engine fuel carrying lines.
Beneath the cabin floor, from FS-335 forward encompasses the cargo compartment and spaces containing some electrical equipment, including the inverters. Sufficient unburned cargo and airplane material was recovered from this area to eliminate the possibility of inflight fire in this portion of the fuselage. Likewise, the inverters did not reveal any signs of operational over-temperature. The investigation included a check of the cargo and stowed personal luggage for hazardous materials with all indications being that none were aboard.
The remaining portion of the fuselage beneath the cabin floor between FS-335 and FS-414 is known as the electrical bay. The generator and starter circuits are brought into the fuselage by means of feed-through studs through the fuselage skin at each side of this compartment. Through stud detail design provides specific safeguards to preclude conductor to ground faults. Circuits from these studs run to main distribution panels mounted on the forward face of the main spar. Also within this compartment are most of the electrical system control and switching relays utilized in the electrical supply system, the batteries, and a freon compressor. Hydraulic lines, some of which may be pressurized, depending on system demand, also run through this compartment. A degree of isolation of the hydraulic lines from the electrical components was attained in the design by physical separation and barriers. Fire damage in the electrical bay area was extensive, particularly on the left side. The majority of the components were destroyed or not recovered. Examination of the recovered electrical components did not reveal any evidence of a heat generating fault. One battery terminal did show an arc produced mark. No evidence of a hydraulic line leak was found. There was no consistent pattern of inflight fire discernible in this area. Smoke patterns on the main spar cap as well as soot and discoloration patterns on seat track pieces that were installed between FS-335 and FS-414 were given detailed attention. Clean fracture and scrape marks next to sooted or discolored areas and discoloration of seat track pieces on the underside, which is exposed in the electric bay, contrasted with clean upper surfaces which are in the passenger cabin but not exposed.
An inflight fire existed in the passenger-occupied portion of the cabin. The only flammable liquid carried as a part of the airplane above the fuselage floor is hydraulic fluid in a reservoir located in a compartment between the carry-on luggage rack and the lavatory. The reservoir was damaged by impact and fire and was empty. The fire damage pattern in and about the reservoir compartment did not support hydraulic fluid as a contributing factor to the fire. Another source of flammable fluid known to have been aboard the aircraft was a one-gallon can containing a commercial paint modifier. This can was recovered in the wreckage area,crushed with no evidence of fire damage to either the can or its paper wrapping.
Fire damage and smoke patterns were found in three general areas of the passenger cabin. These were (1) Beneath No. 4 window on the left side from approximately FS-388 to FS-395; (2) On the cabin side of the forward bulkhead, FS-198, on its left side; (3) On the exterior wall of the forward lavatory, right side at FS-232. There were other isolated areas of fire or soot damage throughout the cabin including the lower half of the No. 9 window, on several public address system speakers, and on some passenger seats.
Under the No. 4 window there was heavy sooting of the shear cleats of the stringer which were tightly compressed against the fuselage skin during breakup. Clean (unsooted) breaks were noted in the sooted areas under the window as well as clean areas where flush rivets had been pulled from the structure.
Two large sections of the left forward bulkhead, FS-198, had received fire damage. The vinyl material that covers the top half was missing but the backing material was not damaged while the bottom half was heavily matted with a deep soot pattern. Another piece of decorative wall material was heavily sooted and matted but the splintered edges of the plywood backing were comparatively clean. The front wall attachment bracket for the blanket rack was moderately sooted but was clean under the rotated reinforcement plate. Most of the fuselage former sections from FS-399 aft to FS-618 displayed distinct unsooted areas that were covered by blanket rack support brackets before breakup. A portion of a soundproof window was recovered coated on the inside with soot and a white deposit which was similar to the white smoke given off when the vinyl cover on the left forward bulkhead is burned. The plastic window material had flow marks on it which indicated that it had been heated to approximately 626°F., while in an upright position. There was also evidence of pre-impact fire in the carry-on baggage rack and other areas in the forward cabin.
The firefighting equipment aboard the aircraft was examined to see if it had been used. Fire extinguishing systems are provided for the engines, underfloor cargo compartments, the Janitrol heater, and the cockpit and cabin. Three of the four engine bottle discharge heads were recovered but none showed signs of having been discharged by electrical means. The forward cargo compartment CO2 extinguisher had been fired electrically. No positive determination could be made as to whether the Janitrol heater compartment CO2 extinguishers had been fired. One of the two cabin CO2 fire extinguishers had been discharged and a cabin water extinguisher had been prepared for discharge; however, it had not been expended.
A flight crew walk-around oxygen bottle was recovered with the control valve open. The rubber diaphragm in the regulator had been discolored by smoke. One passenger oxygen bottle was found with the shutoff valve "open." a portion of one of the three installed flight crew full-face smoke masks was recovered but there was no evidence of its having been in use.
All three supercharger spill valve actuators were found with their actuating rods in the "supercharge spill" position. This setting would dump supercharger air overboard rather than using it to pressurize the cabin. The cabin pressure control located in the cockpit was set at the "sea level" selection and the cabin outflow valves were found in the manually depressurized position. The unpressurized flight valve electrical actuator was found in the "depressurized" position.
The interior locking mechanisms for the No. 4 and No. 9 windows on the left side were found in the unlocked position, and the captain's direct vision (DV) window was found unlocked and partially open. The copilot's DV window track and frame with attached cockpit liner showed evidence of heating and sooting. Adjacent portions of the cockpit liner that had been covered before breakup were clean. The window position at impact could not be determined. There was no evidence of an inflight fire originating in the cockpit portion of the fuselage.
The engines, underfloor cargo compartments, and the Janitrol heater are equipped with fire detection systems. The captain of the previous flight in this aircraft testified that he detected no problems with the fire warning systems when he tested them.
There is no smoke detection system other than the crew sense of smell or observation of smoke in the aircraft.
In normal operation cabin air is drawn down under the cabin floor and circulated back into the cabin by the recirculating fan through ventilators in the cockpit and cabin. Any smoke generated under the cabin floor would be transferred to the cabin and cockpit within seconds where it would be seen or smelled by the crew and passengers.
This accident was non-survivable and no studies were made of the structure from that standpoint.
The passenger that fell from the aircraft exited through an emergency window over the left wing. Witnesses who saw him fall said he did not strike the empennage but fell nearly straight down. An aerodynamic study as well as the body injury pattern confirmed the witness observations. He died of injuries received on impact with the ground. He had received burns on the hands, face and neck before death but had only a few carbon particles in his trachea and a carbon monoxide level of five percent in his blood. The upper portions of his clothes were impregnated with soot. There was no evidence of ground fire where the body was recovered.
The Armed Forces Institute of Pathology (AFIP) performed a number of tests of specimens from both crewmembers and passengers for the Board. Tests for carbon monoxide were not done on the flight crew due to a lack of suitable specimens. Passenger toxicological examination results were negative; no elevated carbon monoxide levels were found; no significant amount of alcohol was found; and tests for methylbromide yielded negative results. Histological examination of the seven recovered respiratory tract specimens revealed only a small number of carbon particles in each.
Federal Bureau of Investigation laboratory tests were accomplished to verify evidence of heat and smoke deposits; analyze various deposits on aircraft parts and on the free-fall victim's clothing, and evaluate possible evidence of sabotage. No residues were found to indicate that an explosion occurred aboard the aircraft. The reports did describe evidence of considerable heat or fire damage sooting to various components within the cabin, cockpit, and under the cabin floor aft of the cargo pit and forward of the main spar.
With the Board's concurrence, British Aircraft Corporation laboratory tests were conducted on samples of aircraft structure, components, and the free-all victim's clothing. These examinations consisted of exposure of ultra-violet light, X-ray, infra-red and emission spectroscopy, microscopic and visual observation. Their test results indicated that there was evidence of cabin fire on seats, windows, forward bulkhead trim, and the carry-on luggage rack. The sidewalls of the carry-on luggage rack had been exposed to temperatures on the order of 122°F., and the plastic material from a soundproof window had been exposed to temperatures of 620°F. They were best able to duplicate the heat damage to the free-fall victim's clothes by burning and quenching a fire of lighter fluid in a sample of the material. They did note the presence of black deposits on the bottom of seat track samples, taken from an area over the electrical bay, with none on the top of the samples. They concluded, however, that the deposits had more the characteristics of lacquer than soot.
The National Bureau of Standards prepared a number of electron photomicrographs of carbon specimens taken from the free-fall window and passenger's clothing, and from various components of the aircraft found at the main wreckage site, primarily from under the cabin floor between FS-317 and FS-414. They conducted tests to determine the effects of various temperatures on paint on pieces of seat track and underfloor runners taken from the aircraft wreckage. They also examined the Janitrol fire extinguisher bottle firing strip. Aluminum paint samples exposed to heat of less than 400°F. for two minutes showed no visible effects. Color changes began after exposure for two minutes at 400°F. and blistering began in two minutes at 450°F. At 900°F. the paint darkened to dark brown or black. These results were used in conjunction with the electron photomicrographs to study various components of the aircraft for evidence of fire in flight. The examination of the Janitrol bottle firing strap was inconclusive and no determinations could be made regarding the conditions under which it fractured.
The photomicrographs of the carbon specimens were forwarded to a specialist in an effort to determine the identification of the material that produced the soot found on the wreckage and the free-fall victim's clothing. The carbon deposits taken from both the free-fall items and under floor wreckage at the crash site were identified as being the produce of incomplete combustion of saturated aliphatic hydrocarbons. Examples of this type of fuel are kerosene, gasoline, paraffin, hydraulic fluid, lighter fluid, and naphtha. Of these examples, only kerosene, hydraulic fluid, and lighter fluid were known to be aboard the aircraft. The lighter fluid was not known to be aboard in sufficient quantity to produce the amount of fire experienced. Kerosene in the form of engine fuel and hydraulic fluid were aboard the aircraft in quantity.
Additional studies of the fire damage were made by a specialist from the Massachusetts Institute of Technology and tests to determine the effect of heat on aluminum surfaces were made under his direction. These tests, made under laboratory conditions, included the exposure of numerous painted and unpainted specimens to open flames in one series and to oven heating in another series. In the open flame tests, kerosene, hydraulic fluid and turbine engine oil were used with both oxidizing and reducing atmospheres to produce soot deposits on the test specimens. In these tests it was found that soot was deposited only on those portions of specimens in the line of flame impingement. Scratches in areas of light sooting remained clean and bright, but became dark with progressively heavier deposits of soot. The soot build-up in scratches and cuts was also a function of flame impingement angle. In another series of tests, individual specimens were dipped in one of the above-mentioned fluids prior to insertion in the oven. They were retained int he oven for five minutes at temperatures ranging from 450°F., to 1000°F. Both oxidizing and reducing atmospheres were used. In these tests painted aluminum surfaces darkened progressively to 800°F. Above this temperature the dark coloring disappeared and the surfaces became silver gray. Between 700° and 800° the fluid deposits burned, leaving black deposits on all surfaces. At lower temperatures scratches and edge cuts on specimens remained bright, independent of the surface discoloration caused by the fluids and other deposits on surface finishes on the specimens. From these tests and comparison with the underfloor wreckage of N7405, he concluded that there was no positive evidence of an underfloor fire prior to impact.
Investigation of insurance purchased by passengers and crewmembers disclosed no suspect areas. Nothing unusual was noted regarding the passengers or baggage that went aboard the aircraft and there was no known hazardous cargo aboard the aircraft.
Crew training and certification were not considered to have a causal relationship to this accident. The weight and c.g. of the aircraft was within limits at takeoff and computed to have stayed within limits until the crash. There was no evidence that the weather, sabotage by explosion, any failure of the powerplants, or the primary aircraft structure played a part in this accident. The Air Traffic Control handling of the aircraft was routine throughout the flight. There is no evidence of improper aircraft maintenance or that the aircraft was not airworthy at the time of its departure from Washington.
There was no indication of any difficulties aboard the aircraft until after it passed Holston Mountain VOR. Having passed the VOR, the crew began a normal en route descent in VFR conditions that would have brought them into the Knoxville area at a reasonable altitude to maneuver for a landing. Their descent was probably normal, i.e., approximately 1,000 feet per minute until they canceled their IFR flight plan at 1802:45. There was nothing in their transmissions to indicate they were having any difficulty at that time. The aircraft should have been at approximately 11,000 feet and about 24 miles southwest of Holston Mountain when they canceled their IFR flight plan.
At some time during the descent, the aircraft deviated to the south of V16 but was proceeding approximately parallel to the airway. No reason can be assigned to this deviation. The first witness believed to have seen the aircraft was 38 nautical miles southwest of Holston Mountain. He estimated the aircraft to be 4,000 feet (approximately 5,500–6,000 m.s.l.) above the terrain and the aircraft appeared to be normal at this time. The aircraft appeared to be following a nearby river, in a descent, and was about 8 miles south of the airway centerline. Three miles farther along the flightpath, at approximately 1810, the aircraft was observed about 500 feet above the ground. The aircraft continued to operate at very low altitudes and well to the left of the airway from this point on to the crash. The average rate of descent from initiation to level off at an estimated 500 feet above the ground was about 1,200 ft./min. and the average ground speed was 174 knots from initiation of the descent to impact. This indicated that the flight's airspeed was reduced from a cruising speed of 247 knots to some lower value and that the descent was continued to an altitude above the ground lower than that normally utilized in transport operation.
It is believed that the crew discovered a fire sometime during the period between cancelling their IFR and before being observed in a descent about 4,000 feet above the ground.
As previously stated there was extensive fire damage in the electrical bay. However, this fact alone cannot be considered significant. This area in the Viscount, as in the majority of low wing configured aircraft, is in close proximity to and between the fuel tanks. Thus, in a breakup, this is a likely area to receive a substantial quantity of the spilled fuel and in turn to be heavily damaged by post impact fire. This fire damage pattern has been observed in many accidents where post impact fire occurred. The somewhat conflicting soot and discoloration patterns observed on certain isolated pieces from the electric bay area dictated further considerations with respect to inflight fire. The only likely source of overtemperature in this compartment is a gross electrical fault to ground. The emergency procedure executed by the crew does not support a gross electrical system malfunction. An electrical source smoke or fire emergency is combated by turning the emergency power switch on and placing the battery master switch and generators off. Equipment that was operating at impact and DME operation to five miles before impact shows this particular emergency procedure had not been executed. Historically, under-the-floor fires that have persisted to a catastrophic stage have burned through the relatively light fuselage belly skin, have been observed by witnesses when present, and have left a path of partially burned debris on the ground. This did not happen in this case. Finally, to involve the hydraulic fluid in an electric bay fire would have required two essentially simultaneous failures, fluid leakage and an electrically induced overtemperature or sparking situation for ignition. Physical evidence fails to support either of these occurrences. Although the object of a great amount of investigative effort, the origin of the fuselage fire in the electrical bay could not be established on the basis of the available evidence.
Burns on the free-fall victim and fire-damaged passenger cabin material found remotely from the primary impact and ground fire area established conclusively that there was an inflight fire in the passenger cabin. Evidence of use of the portable cabin CO2 extinguisher and the attempt to use the portable water extinguisher, together with the open valve of a flight crew walk-around oxygen bottle are suggestive of the first officer having gone back to the cabin to fight the fire a few minutes before the crash. Opening the outflow valves, the left side cockpit window, and emergency exits was probably done with smoke evacuation efforts.
Spill valves to "spill" and discharge of CO2 into the baggage compartment are procedural items to combat a cargo compartment fire and were accomplished. It is recognized that accomplishment of these two items is not compatible with the conclusion that an inflight fire did not originate beneath the cabin floor. Any attempted explanation must of necessity be conjecture. However, it is considered likely that as the situation aboard the aircraft became very grave, precise checklist items were supplemented by any action that offered even a remote possibility of being helpful.
The combustible material and source of ignition that started the fire are not known. Although attempts to determine if any passenger had carried any hazardous material aboard the aircraft did not reveal this had occurred, the possibility cannot be ruled out. Such a material, either innocently or with malicious intent may have been in the possession of a passenger. Leakage or spillage of a flammable fluid with accidental or intentional ignition is a possible situation. Substantial aircraft cabin fires are such a rare occurrence that a most unusual and possibly not readily conceivable circumstance is visualized in this instance.
In examining the final maneuver and crash, it is apparent that the aircraft was not under control of the crew. There are a number of hypotheses that can be advanced to explain this loss of control including: distraction of the pilot; failure of the flight control rods due to fire damage; incapacitation of the pilot by heat and/or smoke; a shift of load caused by the passengers moving to the after end of the cabin; an overt act by some person aboard the aircraft, or any combination of these.
There is no probative evidence available to the Board on which to base a determination as to the cause of the final maneuver.
- The flight crew and stewardesses were properly qualified and certificated.
- The weight and c.g. of the aircraft were within limits at takeoff from Washington, and on the basis of known facts, computed to be within limits at the time of the crash.
- Weather was not considered to be a factor in the cause of this accident.
- There was no powerplant or airframe failure prior to the accident.
- There were no known aircrew errors.
- There was an inflight fire in the passenger cabin.
- Firefighting and smoke evacuation procedures were carried out by the crew.
- The free-fall victim was exposed to high heat and heavy soot before he left the aircraft through the no. 4 window. He did not strike the tail, but received fatal injuries due to impact with the trees and ground.
- Sooting by inflight fire was caused by incomplete combustion of an aliphatic hydrocarbon fuel.
- The Board is unable to identify the source of fuel, the ignition point of the fire, or the cause of the final maneuver.
The Board determines that the probable cause of this accident was an uncontrollable inflight fire of undetermined origin, in the fuselage, which resulted in a loss of control of the aircraft.
Copies of recommendations with the corrective action taken are attached. Attachment No. 1 deals with the installed flight recorder and Attachment No. 2 the forward cargo compartment fire extinguisher.
BY THE CIVIL AERONAUTICS BOARD:
- All times herein are eastern standard time based on the 24-hour clock.
- A second stewardess boarded the aircraft at Washington for the trip to Huntsville.
- Holston Mountain VOR is located 92 nautical miles east-northeast of Knoxville VORTAC on Airway V16.
- The copilot's voice was identified as making the radio transmissions after passing Holston Mountain.
- 83° 03' 41" west – 36° 01' 36" north.
- Standard Kerosene turbine engine fuel.
- Primary radar targets are reflections from the aircraft surfaces. Secondary radar targets are electronic returns from a radar transponder aboard the aircraft.
- Fuselage stations (FS) are measured in inches from a zero datum point at the nose of the aircraft.
- The soundproofing window is a second piece of plexiglass mounted inboard of the regular window covering the entire window opening. These soundproof windows are located at the No. 1 and 2 windows positions and can be identified by their lack of contour and the special plastic used in them.
- Any concentration of less than 10 percent carbon monoxide is considered negative.