Accident at Near Cairo, Egypt, 31 Aug 1950

Accident narrative

On August 31, 1950, Trans World Airlines Flight 903, a Lockheed Constellation, experienced a failure of the No. 3 engine and a subsequent fire near Cairo, Egypt. The crew turned the aircraft back toward Cairo, but the burning engine fell free, and an attempted landing on the desert resulted in the destruction of the aircraft. All 48 passengers and 7 crew members were killed.

**The Flight**

Flight 903 originated in Bombay, India, on August 30, destined for New York with several intermediate stops. The aircraft arrived in Cairo at 2217, where a complete crew change was made. The new crew included Captain Walton B. Webb, who had accumulated 10,664 flying hours, and First Officer H. J. Hammitt, who had 6,355 hours.

The aircraft was fueled to a total of 3,500 gallons. At takeoff, its weight was 99,557 pounds, well below the maximum allowable weight of 107,000 pounds, and its center of gravity was within specified limits. Weather over the route was good. The flight departed Farouk Airport at 2335, planning to cruise at an altitude of 14,500 feet.

At 2355, the Cairo control tower received a routine position report from the flight: “off Cairo 2335 25 miles out of Cairo at 2343 estimated time of arrival Rome 0530.” The tower acknowledged the message, but the flight did not reply with the customary acknowledgment. No further messages were received from the aircraft.

**The Accident Sequence**

Around the time of the position report, persons on the desert northwest of Cairo saw the aircraft afire in flight. The aircraft was approximately 67 miles from Cairo, on course, at an altitude of approximately 10,000 feet.

Based on witness statements and wreckage distribution, the Board reconstructed the flight path. After the fire was first seen, the aircraft continued on course for about 10-1/2 miles. Still burning, it then turned to the left and descended, heading back in the general direction of Cairo. Approximately 13-1/2 miles along this reverse course, the burning engine fell to the desert. The flight reduced its rate of descent, assumed a heading of about 210 degrees, and held this heading until it struck the ground in a relatively flat position about 52 miles northwest of Farouk Airport.

Stopped personal timepieces established that the crash occurred at 0003.

**What the Investigation Found**

The wreckage was sighted the following morning by another TWA aircraft. The Constellation had struck the ground under control and nearly level laterally, with the right wing slightly depressed and a slight nose-down attitude. The landing gear and flaps were retracted. The aircraft disintegrated over a fan-shaped area stretching about 700 feet, and the bursting of the wing fuel tanks resulted in a general fire. Autopsies of the captain and copilot disclosed that neither had breathed flame or hot gas, indicating there had been no fire in the cockpit prior to the crash.

Investigators found the emergency fuel shut-off valve for the No. 3 engine in the closed position. Four of the aircraft's six fire extinguishing bottles had been discharged prior to impact. The aircraft's control, electrical, and radio systems showed no evidence of malfunctioning prior to the crash, with the exception of a probable loss of aileron boost. Engines 1, 2, and 4 were developing power with their propellers in forward pitch at the time of the crash.

The No. 3 engine, with its propeller and engine mount, was recovered back along the flight path at a distance of about 3 5 miles from the impact site. The propeller was within 3-1/2 degrees of the full feathering position. Between the No. 3 engine and the main wreckage, investigators recovered many pieces of molten metal identified as coming from the top surface of the wings.

Teardown of the No. 3 engine revealed the main failure was the rear master rod bearing. As the engine continued operating, the rear crank pin overheated and collapsed until there was approximately an inch difference in the diameters of the bearing and the crank pin. This allowed the effective piston stroke to increase until the pistons struck the valves and cylinder heads, causing all rear articulated rods and the rear master rod to fail.

These flailing rods sliced through the walls of the rear cylinders and crankcase, displacing and distorting the inter-zone baffle immediately aft of the cylinders. This destroyed the baffle's function as a fire seal. Severed lines liberated lubricating oil and fumes, resulting in a fire that was relatively light in Zone 1 (forward of the baffle) but so severe in Zone 2 (aft of the baffle) that it melted the adjoining dural structure rearward of the firewall in Zone 3, allowing the engine to fall free. The fire then continued through Zone 3 and the adjacent wing structure.

The engine had operated approximately 1,100 hours since installation; the specified overhaul period was 1,200 hours. The oil had not been changed, but merely added to as required, which was company procedure at the time. Maintenance records showed that 18 hours before the accident, the No. 3 engine experienced a temporary, unusually high difference of approximately 40 degrees C between the in-flowing and out-flowing oil temperatures. The operator did not consider this indicative of potential trouble, as it was within their arbitrarily set safe spread of 45 degrees C, and the in-flowing oil had not exceeded the manufacturer's maximum of 83 degrees C.

The Board evaluated three possibilities for what initiated the master rod bearing failure: 1. A burned No. 9 piston was found, but it is believed such a failure would have been evident by a loss of power and oil before initiating a master rod bearing failure. 2. The Board noted a slight possibility that a balancer seized during the high oil temperature event 18 hours prior, introducing metallic parts into the lubricating system. 3. The engine exhibited a normal accumulation of hard, caked sludge in the hollow crank pins. A study of sludge accretion in this model engine disclosed that after about 700 hours, the sludge level reaches the tops of the oil feed tubes. While sludge normally remains solid, it may shift and obstruct oil flow. Because the engine had been in operation for 1,100 hours, the Board noted a distinct possibility that sludge obstruction of oil flow was the primary cause of the failure. The investigation eliminated oil dilution as an initiating factor in loosening the sludge, as the aircraft had not been used in cold weather since June 1950.

Following this and previous master rod bearing failures, TWA discontinued the practice of not changing oil between overhauls, instituting changes at periods not exceeding 400 hours. The carrier also discontinued oil dilution in all similar engines and issued a bulletin to flight crews detailing the early symptoms of master rod bearing failures. The bulletin warned that increased oil flow through a failing bearing would rapidly indicate as a drop in oil pressure and an increase in oil temperature, and urged crews to feather the engine when these symptoms were observed. A review of six previous failures showed they were recognized for an average of 12 minutes before feathering took place.

**Probable Cause**

The Board determines that the probable cause of this accident was failure of the rear row master rod bearing causing an uncontrolled fire which precipitated a crash landing.