Accident Overview

Illustration of DC-10 Engine Locations
Illustration of DC-10 Engine Locations
(View Large Image)

Graphic of cutaway view of the CF6-6 engine
Graphic of cutaway view of the CF6-6 engine
( View Large Image)

History of Flight

On November 3, 1973, while National Airlines DC-10 Flight 27 from Miami to San Francisco was cruising at 39,000 feet, the number 3 engine fan assembly completely disintegrated.

The number 3 inlet and fan case separated simultaneously with fan blade liberation, allowing the blades to impact the airplane at high speed. The liberated blades struck the fuselage, right wing, engines number 1 and 2, and caused a cabin window to fail. This resulted in the expulsion of one passenger through the opening during rapid decompression. Many passenger oxygen masks failed to deploy, causing a panic among the remaining passengers. The engine debris also caused damage to two of the aircraft's three hydraulic systems resulting in loss of all hydraulic fluid for both systems.

At the time of the number 3 fan failure, the pilot and flight engineer were "experimenting" with the relationship between the engine's fan speed (N1) signals and the automatic flight control systems. The pilot and flight engineer appeared to be curious as to where the autopilot/auto throttle received its engine speed signal (i.e., N1 gauge or N1 tachometer generator). The crew reported that their "experiment" was conducted as follows:

  • Autopilot selected in airspeed mode - 257 Knots-Indicated Air Speed (KIAS)
  • Flight engineer pulled circuit breakers for N1 tachometers on engine numbers 1, 2, and 3
  • Target airspeed reduced by five knots
  • When throttles retarded, pilot disengaged auto throttle
  • Crew then heard explosion and cabin decompression


Photo of damage to the aircraft Photo of engine no.3 throught missing cabin window
Photos of damage to aircraft
( View Large Images)

The crew declared an emergency and executed an emergency descent. There was an uneventful landing at Albuquerque, New Mexico.

The oil tank in the number 1 engine had been penetrated by fan blade debris and upon landing had lost all oil, and the number 1 engine constant speed drive generator line was severed. Engine number 2 ingested fan debris and sustained minor foreign object damage. The failure of a substantial number of oxygen masks to deploy appeared to result from multiple electrical failures.

Analysis of cockpit voice recorder data shows that the number 3 engine accelerated from 97% N1 to 100% N1, fluctuated for a few seconds and failed at 100%. The flight data recorder was inoperative. Engines number 1 and 2 also accelerated from 97% N1 to a maximum of 107% N1.

Photo of No. 1 engine cowl showing impact damage
Photo of No. 1 engine cowl showing impact damage

The photo to the right is the number 1 engine cowl showing impact damage from the number 3 engine debris. Engine debris penetrated the oil tank resulting in total loss of engine oil upon landing.

Investigation

The initial triggering event was not completely understood during the course of the investigation, due to the complex nature of the vibratory phenomena later identified. The investigation centered on vibratory characteristics of the fan blades and fan case. All parties did not agree on the initiating event. The question remained, "Did the inlet failure cause tip rub?" or "Did tip rub cause the inlet failure?" There was general agreement that the failure mechanism somehow involved a fan case 6th order wave mode traveling at one half the fan speed in the direction of fan rotation.

It was also agreed that the fan blade assembly fore and aft rocking vibratory mode was a contributor to the failure mechanism. This rocking mode also involved a 6th order traveling wave shape at one half fan speed in an opposite direction to the fan.

Illustration of Campbell Diagram
Illustration of Campbell Diagram

The investigators concluded that both conditions had to be present for a massive engine failure such as this to result:

  1. A coincident and coupled fan case wave and blade "rocking" wave mode, and
  2. Fan blade tip rub.

Note: All large, high bypass, fan engines commonly experience fan tip rubs, especially early in service. High "g" aircraft maneuvers, high physical fan speeds, and other operating conditions where tip clearances are reduced due to component deflections can result in minor abrasion of tip shroud material, or "tip rubs". This "wearing in" of fan and inlet components is an expected condition, and is intended, in order to establish "final" tip clearances between the blades and fan case. This clearance is also very significant in establishing installed engine thrust.  

The concept of sympathetic vibratory modes is illustrated in this diagram, known as a Campbell Diagram, showing the vibratory frequencies of various components plotted in overlay - in this case, a fan case and fan blades, and their vibratory modes up to 4th order. The components have various vibratory frequencies that are generally independent of each other. However, where the lines intersect, the modes are sympathetic, and if fan blades and fan case come in contact, a divergent feedback process results in a catastrophic energy release. In this accident, the fan blades and fan case were vibrating in sympathy with one another, each at 6th order, and their contact caused the destruction of the forward section of the engine.

View CF-6 Vibratory Characteristics Flash Animation.

Photo of National DC-10 safe landing after incident
Photo of National DC-10 safe landing after incident

Although the flight crew was experimenting with the aircraft systems and not following an approved procedure, the root cause of the failure was not the improper flight crew actions in the cockpit, but a design problem with the engine fan, fan case and inlet combination.

During this accident, the maximum speed of the fan rotor on the number 3 engine when it failed was determined to be approximately 100% which is below the maximum red line speed of 111%. Had the number 3 engine not failed on the National Airlines flight, it was believed that eventually a similar vibratory condition would have been experienced in service with similar catastrophic results.

Back to top