NTSB Factual Report

On December 25, 2009, approximately 2000 central standard time, an Aerospatiale ATR 72-212A airplane, N494AE, operated by Executive Airlines, Inc. as Eagle Flight 4756, experienced a flight control system malfunction while en route to Dallas/Fort Worth International Airport (KDFW) Dallas, Texas. The two flight crew members, two flight attendants, and 41 passengers were not injured. The domestic passenger flight was being operated under the provisions of 14 Code of Federal Regulations Part 121, on an instrument flight rules flight plan. Night visual meteorological conditions prevailed at the time of the incident. The flight originated from the Midland International Airport (KMAF), Midland, Texas, at 1827.

While in level flight, with the autopilot on, the flight crew received a pitch mistrim message. In accordance with check list procedures they disconnected the autopilot and discovered that fore and aft movement of both control columns was stiff with limited travel and did not feel normal. They could move the elevator controls a maximum 1 inch pitch up and pitch down. The flight crew twice attempted the Jammed Elevator procedure in an effort to uncouple the elevators. Despite their attempts they did not succeed in uncoupling the elevators. While coordinating with their company’s maintenance operational control they slowed the airplane to 180 knots and found that they had regained increased control of the elevator.

On a final approach and after configuring for landing, the elevators again felt jammed. The flight crew conducted a go-around and declared an emergency. During the second landing attempt the flight crew still had both control columns partially jammed. They performed a shallow approach to a smooth landing. While taxing to the gate, they noted the elevators were still jammed. After parking, with engines shutdown, the flight crew regained full control of elevator.

An after landing inspection by company maintenance personnel found the brackets holding the left elevator down limit stop had fractured and separated from the hinge fittings and the separated down limit stop was restricting movement of the left elevator. The left elevator hinge assembly and pitch uncoupling mechanism (PUM) disconnect switch were removed from the airplane and sent to the National Transportation Safety Board Materials Laboratory for further examination.

The inboard and outboard L-brackets that attached the lower stop to the hinge fitting were fractured. The inboard L-bracket was fractured at the bend. The outboard bracket was rotated approximately 90 degrees with respect to the stop and had fractured near the nut plate.

The lower stop shim stack was subjected to deformation and tearing. The shim stack was composed of four layers. Starting from the face of the stack in contact with the hinge fitting / L-brackets and progressing toward the stop, the layer sequence was laminated shims, solid shim, laminated shims, and laminated shims. The total shim stack thickness was 0.15 inch (3.8 mm). The maximum allowed shim stack thickness was 0.062 inch (1.6 mm). The shim stack specification called for a solid shim 0.030 inch with a 0.032 inch maximum stack of laminated shims. A section of the laminated shim layer in contact with the hinge fitting was missing. The adjacent solid shim was deformed and torn. The laminates in the two remaining layers were deformed and layers were splayed apart between the brackets.

Features on the inboard L-bracket fracture surface were consistent with high-cycle fatigue initiating from the inner bend and progressing to the outer bend. While much of the fracture surface was damaged or covered with debris, isolated regions showed features consistent with fatigue striations.

Higher magnification examination illustrated curved striations consistent with fatigue. The curvature of the striations was consistent with the crack originating at the inner bend. The size of the fatigue region and the direction of crack propagation were consistent with high-cycle unidirectional bending fatigue.

The positioning of the L-brackets with respect to the support flanges on the hinge fitting was examined by reattaching the L-brackets with clekos and using a straight edge to inspect for gaps. The inboard bracket was slightly proud of the hinge fitting resulting in a gap between the straight edge and the support flange. The outboard lower stop bracket did not protrude below the support flanges. The alignment of the inboard bracket could not be reliably determined because of the fracture.

The PUM disconnect switch was stuck in the closed position when received by the Materials Laboratory. The closed position is the normal state of the switch when the left and right elevators are coupled. When the PUM is disconnected, the elevators are uncoupled, the switch opens, and a cockpit alert indicator is illuminated.

The continuity of the PUM disconnect switch leads were tested with a multimeter. The six switch leads were labeled alphabetically from A to F. Electrical continuity was detected between leads B and C and between leads E and F. This pattern was consistent with the switch being stuck in the closed position. The unit was not further examined and the reason for the switch failure could not be determined.

Normal procedures in the flight manual for the preliminary cockpit preparation and parking require confirmation that the gust lock is engaged. The procedures specify that the gust lock must be released during the before takeoff checks and the gust lock must be engaged after landing. The operator’s procedures caution that the gust lock must be engaged when the airplane is moving on the surface or when parked unattended. There are additional instructions to confirm that the gust lock is engaged prior to taxi or when leaving the aircraft.

A review of data from the flight data recorder (FDR) from the incident airplane showed several events when the elevators were in a position consistent with the gust lock being properly engaged before takeoff and after landing. Of the 92 total flights recorded, the FDR data showed the elevators were in a position consistent with the gust lock being properly engaged before takeoff and after landing on about 9 of those flights. The FDR data also showed numerous events, including taxi time, on the remaining flights when the elevators were in contact with the lower stops. This position is consistent with the gust lock not being properly engaged. Therefore the elevators were allowed to slam against the lower stops. This condition would allow the generation of repeated stress on the elevator stops and the brackets holding those stops.

On July 30, 2010, the operator issued a Safety Alert Bulletin Alert 2010-005 “Topic: Elevator Gust Lock – Flight Crews and Mechanics” reminding them of the operator’s gust lock engagement procedures.

On July 1, 2010, the European Safety Agency (EASA) issued Airworthiness Directive AD 2010-0138 mandating an inspection of all ATR-42 and ATR-72 airplanes in the worldwide fleet.

Effective on December 15, 2011, the Federal Aviation Administration (FAA) issued Airworthiness Directive FAA-2011-0721 mandating an inspection of all ATR-42 and ATR-72 airplanes.

NTSB Probable Cause

A fatigue failure of an elevator stop bracket, which resulted in a partial jam of the elevator control. Contributing to the failure was the improper installation of shim stacks, poor alignment of the brackets, and cyclic stresses acting on the lower stop, generated by the repeated improper use of the gust lock system.