Office of The Inspector General USAF
Norton Air Force Base
San Bernardino, California

19 September 1952

Report of Special Investigation of Major Aircraft Accident
Involving B-36D, SN 49-2661, at San Diego Bay,
San Diego, California, on 5 August 1952


1. B-36D aircraft, SN 49-2661, on bailment to Consolidated Vultee Aircraft Corporation, (Convair), San Diego, California, crashed into San Diego Bay at 1430 PDT, 5 August 1952 while on a normal shakedown flight following completion of "San-San" project modification. The aircraft was destroyed by impact and explosion. Four of the eight crewmembers received minor injuries, two were uninjured, and two are missing.


2. It is concluded that:

a. The primary cause of the accident was fire in the vicinity of No. 5 engine (see pars 10 and 12).

b. The probable causes of the fire were:

(1) A malfunction of the No. 5 alternator and/or alternator constant speed drive (CSD) resulting in extreme heat and ignition of a magnesium fire (see par 20).

(2) A fuel leak in the vicinity of No. 5 engine (see pars 17, 18 and 19).

(3) Material failure of the No. 5 engine turbo supercharger or exhaust systems, allowing residual oil to become ignited by exhaust flame (see par 14).

c. The use of magnesium alloy materials adjacent to potential fire areas in B-36 aircraft constitutes a hazard (see par 12).

d. The Edison fire detection system is not adequate (see par 21).

e. The right scanner detected smoke coming from the No. 5 engine, but did not report the observation in accordance with the procedure outlined in T.O. 01-5EUC-1 dated 15 March 1952. As a result the fire extinguishing system was not activated (see par 22).

f. The flight test crew did not comply with T.O. 01-5EU-355, dated 21 December 1951, in the operation of the electrical system during the flight (see par 20).

g. The emergency fuel shut-off procedure outlined in T.O. 01-5EUC-1, dated 15 March 1952, is inadequate for combating engine and wing fires (see par 17).

h. The issuance of T.O. 01-5EU-352, dated 31 January 1952, subject: "Sealing of Fuel System Hose Connections - All Series RB-36 and B-36 Aircraft," as a red diagonal instead of a red cross compliance constitutes a serious flight hazard in view of the critical nature of fuel leaks in B-36 aircraft (see par 19).

i. The placard instructions on the flight engineer's panel outlining the procedure to follow in combating engine fires are not in accordance with instructions contained in applicable B-36 technical orders (see par 22).

j. The absence of the right scanner from his duty for approximately 15 minutes delayed detection of the emergency (see par 24).

k. Members of the contractor flight crew were not wearing their life vests at the time the emergency occurred (see par 25).



a. Examine possible means of reducing the hazard resulting from the use of magnesium alloys adjacent to potential fire areas in future USAF aircraft and revise the HIAD accordingly (see par 12).

b. Evaluate fire detection systems installed in all USAF aircraft and develop reliable devices which provide positive warning of wing or engine fires (see par 21).


a. Require that contractor flight test crews are familiar with and comply with applicable technical orders for reporting observations of an emergency nature, for engine shut-down emergency procedures and for combating engine and wing fires in USAF aircraft (see par 22).

b. Require that contractor flight test crews comply with T.O. 01-5EU-355, subject, "Operation of the Electrical System of B-36 Series Aircraft in which Exciter Protection Relays are not Installed" (see par 20).

c. Revise B-36 Operating Instructions to insure that the emergency fuel shut-off procedure will isolate engine or wing fires from the fuel manifold system (see par 17).

d. Require compliance with technical orders involving safety of flight items prior to the next flight of the aircraft involved after receipt of kits (see par 19).

e. Insure that the engine shut-down procedure placard in B-36 and RB-36 series aircraft is in agreement with pertinent technical orders (see par 22).

f. Revise B-36 Operating Instructions to require that left and right scanners' stations are manned at all times during flight (see par 24).

g. Take action to insure that contractor flight test crews wear life vests when engaged in over-water flights in USAF aircraft (see par 25).


5. The recommendations contained in paragraph 3 have been referred by letter to the Commanding General, Air Research and Development Command, for action and reply.

6. The recommendations contained in paragraph 4 have been referred by letter to the Commanding General, Air Materiel Command, for action and reply.


7. B-36D, SN 49-2661, took off from Lindbergh Field at 1039 PDT, 5 August 1952 on a routine shakedown flight following completion of "San-San" modification. The aircraft was operated by a Convair civilian flight test crew. After approximately three hours of flight, a decent was made to 4,000 feet where two normal landing gear extensions and retractions were made. The crew then prepared to accomplish a manual emergency gear extension.

8. The right scanner, a qualified Convair B-36 flight engineer, and the radar technician were selected to manually extend the gear. The right scanner experienced no difficulty in dropping the right gear. However, the radar technician, having less experience in this operation, had difficulty in lowering the left gear. He crawled out of the left wheelwell area and requested the right scanner to accomplish the left gear extension also. The radar technician then proceeded from the left wheelwell to the right wing root area adjacent to the right wheelwell. He announced over the interphone that the right scanner was also lowering the left gear. Shortly thereafter, the right scanner emerged from the left wheelwell area stating that the gear had been extended. The radar technician, in the meantime, had detected gasoline fumes coming from the right wing crawlway access port. When the right scanner returned from the left wheelwell area, he was asked by the radar technician to confirm and investigate the fuel fumes noted. He did this but considered the presence of fumes of no consequence. He explained that the fumes were probably coming from the right wing auxiliary and No. 4 tank vent outlets that are below and aft of the wheelwell.

9. Upon completion of these checks, the aircraft was scheduled to land. Immediately upon return to his station, the right scanner detected smoke coming from the No. 5 engine and advised the engineer on station at the panel to feather No. 5 engine. He then instructed the flight engineer to shut off the oil and fuel to this engine. The flight engineer complied with these instructions. Seconds later there was an explosion in the right wing. The Aircraft Commander, in the copilot's seat, glanced out the window at the right wing area and directed that the aircraft be abandoned. This order was heard and obeyed by all crewmembers; however, the Aircraft Commander did not successfully evacuate the aircraft. The B-36 crashed into the sea about three and one-half miles from shore and exploded on impact. The wreckage sank in approximately
XXX feet of water. Six crewmembers were rescued. Two crewmembers are missing.


10. Interrogation of witnesses disclosed that the aircraft was flying straight and level, with the landing gear extended, at an altitude of approximately 4,000 feet MSL when light grey smoke was first observed coming from the right wing area. At approximately 1426 PDT, an explosion occurred in the vicinity of No. 5 engine at which time black smoke and fire were observed. Immediately thereafter No. 5 engine fell from the aircraft, and intensity of the fire increased. As the aircraft lost altitude, witnesses observed seven crewmembers parachute from the aircraft. Small portions of the aircraft were seen to fall free as it continued its descent. The aircraft turned slowly to the right approximately 200 degrees from its original heading and crashed into the sea in a steep dive. At the time of initial impact, there was an explosion followed by fire that covered an area approximately 200 yards in diameter.

11. The No. 6 engine, two sections of the outer right wing panel between No. 6 engine and the jet pod, a portion of the No. 5 fuel tank, a part of an outboard engine mount from No. 4 engine, small pieces of metal, some electrical wiring, and a portion of one bomb-bay door were recovered. However, examination of these parts failed to reveal the cause of the accident.

12. Magnesium alloy is used in the fabrication of casings around and aft of the engine cooling fan, induction system air duct, alternator housing (bullet), major portions of the nacelle skin, aileron skin, engine cooling air duct (tunnel), wing leading and trailing edge, and integral portions of the alternator. The advisability of using magnesium alloy in the fabrication of those components of the B-36 wing and engine nacelle adjacent to potential fire areas is questionable. In this accident it is suspected that the origin of the fire was in the No. 5 engine nacelle. It seems improbable that the fire started in any other portion of the wing, since the right wing appeared to be normal during the time the landing gear was being manually extended by the crew. At the completion of this operation, the crew noticed gasoline fumes and again checked the interior of the right wing from the wing crawlway access port. A this time, no smoke or fire was observed. The absence of any smoke or fire at this time and the fact that the flight engineer did not receive a fire warning light indication, together with the fact that there was no fuel, oil or other instrument variation, suggests that the fire originated from a magnesium source. This is substantiated by the severity of the fire, the heavy white smoke, and by the Aircraft Commander's immediate decision to abandon the aircraft. The intensity of the fire would be substantially increased when fuel lines severed either by the magnesium fire itself, by the reported initial explosion, or by the No. 5 engine falling from the aircraft.

13. The surviving crewmembers stated that immediately prior to the emergency, all reciprocating engines were operating normally. Power settings were 35 inches manifold pressure, 2200 RPM with turbo master control at zero. The mixture controls were normal. Air plugs of jet engines Nos. 2 and 3 had just been activated to the open position in preparation for airstart. Earlier in the flight, the radar technician was unable to check the optical sight on the bombing-navigational equipment because of aircraft vibration. At that time, propellers were being operated at 1550 RPM; however, a subsequent feathering check of all propellers indicated that vibration was not due to any particular engine or propeller. It was the opinion of the radar technician that most of the vibration was caused by No. 3 engine. With the exception of the left scanner, who visually observed some vibration on No. 3 engine, the remainder of the crew did not detect the vibration or consider it of any consequence.

14. The oil systems of the reciprocating engines operated normally before the emergency. There was no instrument indication of an oil leak. The first reported light grey smoke could have been caused by oil leaking on the hot exhaust manifold of No. 5 engine or by an oil leak in the No. 5 engine turbo supercharger. A review of Unsatisfactory Reports on B-36 aircraft, wherein fire was a possibility, revealed there were a total of 163 reports submitted during the period January through July 1952, concerning electrical components which would have caused or contributed to fire. This study indicates that materiel failure of the turbo supercharger could have been a possibility. There were 507 Unsatisfactory Reports submitted during the same period on specific exhaust system items which may have caused or contributed to fire. However, since a magnesium fire was observed in the vicinity of No. 5 engine, it appears that the fire originated from a source other than the supercharger oil.

15. The hydraulic system was normal, except that sluggish operation was experienced with the nosewheel steering on takeoff. The landing gear extension and retraction operation took longer than normal, and the operation of the bomb-bay doors was sluggish.

16. The shakedown mission being flown required a free-fall operation (manual extension) of the landing gear. There was no malfunction of the landing gear system reported. While the right scanner was in the left wheelwell, the other crewmember, standing by on interphone, detected gasoline fumes coming from the right wing crawlway access port. This fact introduced the possibility of a fuel manifold leak near the landing gear components. In an attempt to determine the possibility of this occurrence, an inspection of another B-36 aircraft was accomplished. It was disclosed that there was marginal clearance between the three-inch fuel manifold and the main landing gear pivot shafts. However, after review of testimony and consideration of all known facts, there appears to be no correlation or connection between the free-falling of the gear and the discovery of the gasoline fumes, except that the airstream configuration would have been changed in the wheelwell area at the time the gear was extended, causing fumes to enter the crawlway access port. The Aircraft Commander was advised of the detection of gasoline fumes; however, he evidenced no concern over this condition.

17. Except for the detection of gasoline fumes during the manual extension of the landing gear, it appears that the fuel system operated satisfactorily up to the time of the emergency. Immediately prior to the emergency, fuel was being delivered to the fuel manifold from Nos. 1, 2, 5, and 6 tanks; booster pumps were "ON." Fuel tank valves on Nos. 3 and 4 tanks were closed, with booster pumps "OFF." The left and right auxiliary fuel tanks were shut off. No bomb-bay fuel tank was installed. When the emergency was declared, the flight engineer closed the No. 5 engine fuel valve and No. 5 fuel tank valve. However, since Nos. 7 and 8 fuel manifold valves were not closed, gasoline under pressure would be present in the fire area. If a fuel line was severed upstream of No. 5 engine fuel valve by the fire, explosion, or at the time No. 5 engine separated from the aircraft, the result would be an uncontrollable gasoline fire. Had the flight engineer closed Nos. 7 and 8 fuel manifold valves, gasoline would not be pumped into the No. 5 nacelle area. This emergency fuel shut-off procedure followed by the crew closely parallels the procedure as outlined in T.O. 01-5EUC-1. The technical order should be revised to include appropriate instructions to isolate specific fuel tank and manifold fuel from a suspected fire area.

18. During the inspection of a B-36 aircraft, it was disclosed that a number of fuel system components are closely grouped in the forward area of the Nos. 2 and 5 nacelles. There are no electrical relays or other electrical components in this immediate area which could produce an exposed electrical spark. This area is not ventilated by ram air circulation; however, it is conceivable that if a fuel leak existed in this area, fuel in vapor form could pass from this area aft to the air plug. If this occurred, the vaporized fuel would pass over hot exhaust ducting and, in all probability, would be ignited causing an immediate gasoline fire with resulting black smoke. Considering all factors and the fact the initial smoke was reported by witnesses to be from white to light grey in color, it would appear that the initial fire originated from a source other than fuel.

19. Technical Order 01-5EU-352, subject, "Sealing of Fuel System Hose Connections - All Series RB-36 and B-36 Aircraft," had not been complied with. This technical order was issued to eliminate a possible fire hazard resulting from leaking of fuel line hose connections; however, it could not be determined wether B-36D, SN 49-2661 developed any fuel leak while in flight. The contents of this technical order would indicate that it is a safety of flight item and should require immediate compliance, rather than be placed on a red diagonal status.

20. There was no indication of a malfunction of the electrical system. Alternators Nos. 2, 3 and 4 were paralleled on the entire bus. No. 5 alternator was excited but in a standby status. No. 5 alternator breaker was in the open position; therefore, no indication of its operation was apparent to the crew. All bus-tie circuit breakers were closed. This was the alternator configuration for the entire flight. This procedure conflicts with the procedures outlined in T.O. 01-5EU-355, 21 December 1951, subject, "Operation of the Electrical System of B-36 Series Aircraft in which Exciter Protection Relays Are Not Installed." This technical order requires that Nos. 2, 4 and 5 alternators be operated in parallel and that alternator No. 3 be operated on an isolated bus supplying the loads to that bus. The technical order applicable to B-36D, 49-2661, further states that Watt-VAR meter readings will be observed at 15 minute intervals. This procedure was not complied with by the flight crew. Line voltage and frequency had been adjusted earlier in the flight by means of a precision voltmeter during test of the K-3 Bombing-Navigation System. The voltage, wattage, and frequency indicator readings, when read, were normal with the exception that wattage of No. 5 alternator was always always "ZERO" since that alternator was on standby for the entire flight. Under the flight conditions existing before and during the time of the emergency, there was no demand for a large amount of electrical current. The UR and accident history of B-36 alternators indicates that an undetected malfunction of No. 5 alternator may have occurred. If No. 5 alternator did fail, the crew would have had no indication since this alternator was on standby. The number of UR's submitted on alternators totals 115 during the period 1 January through July 1952. During the same period, there were 97 UR's submitted concerning the constant speed drive. The following alternator malfunctions could have resulted in a fire:

a. Failure of the speed governing system of the constant speed drive would permit the alternator to rotate at excessive speed. This could result in disintegration of the alternator which creates excessive heat and could result in a magnesium fire of the integral components of the alternator.

b. Failure of an alternator bearing would cause mechanical friction between rotor and stator with the same result as above. With No. 5 alternator in a standby status, the above conditions could occur without being detected from instruments on the flight engineer's panel.

21. Two fire warning lights were found to be inoperative during the flight: No. 2 accessory section and No. 6 power section. However, neither of these two areas can be associated with the cause of the accident. Fire warning lights that were reportedly operational did not indicate the presence of the fire in or around No. 5 engine nacelle. A possible reason for this malfunction may be explained by the conditions necessary to activate the Edison fire warning system. The thermocouple unit must be subjected to an abnormally rapid rise in temperature. If the rate of temperature rise is not rapid enough, the thermocouple unit will not produce sufficient voltage to energize the relay that closes the circuit to the fire warning light. It is believed that the rate of temperature rise of the fire that developed in or around the No. 5 engine nacelle was not rapid enough to activate the fire warning system. The No. 5 alternator is not in the immediate vicinity of any fire detection units. Information obtained in the field and in the testimony of this investigation, indicates a high number of occurrences in which the B-36 fire detection system failed to detect a known fire. Study of the experience with fire detection devices in USAF aircraft emphasizes the inadequacy of present systems to provide reliable indications of fire.

22. The fire extinguishing system was not activated at any time. The system is normally activated as step #5 in the procedure for fighting an engine fire described in T.O. 01-5EUC-1. The flight engineer did not discharge the fire extinguisher in No. 5 engine nacelle. This may have been due to the short time interval between following the commands of the right scanner and the Aircraft Commander's order to bail out. The issue of detailed commands by the right scanner instead of reporting the fire as such, was irregular and did not follow the
engine shutdown procedure in the operating instructions of T.O. 01-5EUC-1, neither did his procedure follow the procedure outlined on the placard beneath the engine selection switches of the fire extinguishing system on the main control panel. The Aircraft Commander was monitoring the interphone during the emergency. However, he did not countermand any instructions given by the right scanner, neither did he advise the flight engineer to follow any particular procedure for engine shutdown. All B-36D and RB-36 aircraft inspected at Convair had the outdated placard posted in the aircraft. It is possible that other B-36 aircraft may be placarded likewise. The placard reads as follows:

"Emergency Operating Instructions

1. Propeller - Feather
2. Mixture Control - Idle Cut-off
3. Engine Fuel and Oil Shut-Off Valves - Close
4. Fire Ext. Eng. Sel. Switch - On (5 Secs.)
5. Heat, Anti-ice, Cooling Switches - Off
6. Pressurization Switch - On Unaffected Side
7. Discharge Switch - Alternator Position
8. Do
NOT Restart Engine"

The procedure, as indicated above is not in agreement with T.O. 01-5EUC-1. It is apparent, from testimony submitted, that Convair civilian flight personnel have not been following the procedure outlined in the technical order. It is imperative that the fire extinguisher be activated when smoke and/or flame is visible.

23. All radio communication equipment, including the interphone, operated normally throughout the flight. Before evacuating the aircraft, the radio operator using the liaison transmitter and receiver, contacted the Convair ground station on 3280 kcs. declaring an emergency. Other testimony indicated that the liaison transmitter meter readings were normal, which corroborates the flight engineer's testimony that the electrical system was operating at the time of the emergency.

24. The flight crew scheduled for this mission consisted of eight personnel. They were: Aircraft Commander, copilot, flight engineer, assistant flight engineer, radio-radar operator (radar technician), assistant radio-radar operator, right scanner, and left scanner. All crewmembers were well qualified to perform their respective duties, with the exception of the radar technician who was not qualified to free-fall the landing gear. Only two crewmembers were in the aft portion of the aircraft. When the right scanner went to the wheelwell areas in order to free-fall the landing gear, his position was unattended for approximately 15 minutes. The left scanner remained in his position prior to the time the emergency was declared. From his position he did not notice any smoke passing by the right scanner's blister; however, after the right scanner returned to his position, the left scanner, when near the right scanner's position, was able to observe light grey smoke coming from the air plugs of No. 5 engine. The left scanner explained that, due to the undercast cloud conditions, it was very difficult to see the smoke and only when he saw it coming from the air plug, could he follow the smoke trail aft.

25. Although Convair standing operating procedures direct that a life vest will be worn when flying over water, only two crewmembers were wearing life vests when they parachuted. Other crewmembers had left their life vests in their A-3 (parachute) bags that were aboard the aircraft. It appears from the testimony that there is a general disregard for the wearing of over-water equipment.

26. There were no discrepancies noted in the aircraft engineering forms maintained by Convair. The company inspection forms that were reviewed contained discrepancies covering the period 3 July 1952 to 4 August 1952. There were a total of 23 discrepancies concerning the right wing area. Of these discrepancies, there were two fuel leak items and three oil leak items. Ten of the 23 items concerned No. 5 engine nacelle and were of a varying nature but were not significant in establishing a pattern. According to testimony, all discrepancies were satisfactorily corrected prior to this flight.


27. The following data pertaining to aircraft accident investigation of B-36D, SN 49-2661 are on file in the Directorate of Flight Safety Research and can be obtained on request:

A. Special Orders Directing the Investigation
B. Western Air Procurement District Investigation Orders
C. Statistics
D. Statements of Witnesses
E. Testimony of Flight Crew to Convair
F. Testimony given to Aircraft Accident Investigation Board
G. Sequence of Events
H. CAA Incident Report (Tower Report)
I. Flight Clearance Form
J. Aircraft Flight Release Form (Form F Data)
K. Discrepancies Written up on B-36D, SN 49-2661
L. Check Sheets, B-36D, SN 49-2661
M. Convair Flight Safety Manual and B-36 Flight Crew Indoctrination
N. Map of Crash Area
O. Statement of Examination of Recovered Wreckage
P. B-36 Aircraft Accident History
Q. Photographs

Brigadier General, USAF
Director, Flight Safety Research