WPR15IA046

The failure of the gust lock actuator solenoid for reasons that could not be determined because postaccident examination of the rudder gust lock system revealed no mechanical malfunctions or failures that would have precluded normal operation.

The airline transport pilot was preparing to take off for a personal cross-country flight. He reported that the preflight preparations were normal, including the removal of the rudder gust lock pin. He stated that the taxi out for takeoff was normal but that he "detected an abnormal 'feel' to the rudders." He added that all of the turns to the runway were right turns and that he "easily executed" the turns. Following rotation on takeoff, the pilot noticed a significant right yaw and roll, which he initially countered with left aileron. As the airplane continued to climb, the airplane's flight was uncoordinated, and the pilot had to continuously apply left rudder to counter the right yaw; he also attempted to apply left yaw trim, but neither action resolved the issue. At this time, the pilot advised air traffic control (ATC) that he needed to return to the departure airport, and ATC cleared the pilot for a visual approach. The pilot landed the airplane; only minor damage to the left wing tip was incurred as the pilot tried to unsuccessfully correct (with the right rudder and aileron) a drift to the right after landing. The initial examination of the airplane revealed that the rudder gust lock had failed in the locked position. Upon further examination, it was determined that the gust lock actuator had likely begun to retract as commanded by the pilot when he removed the rudder gust lock pin during the preflight/startup checks. However, the solenoid located within the gust lock actuator failed, which resulted in the pin no longer being held in the retracted position. Once the pin released from its retracted position, it engaged the ball nut, which then stopped the rotation of the actuator shaft. The failure occurred at a time when the gust lock bellcrank had not fully disengaged from the rudder quadrant, which likely resulted in the rudder being locked throughout the taxi, takeoff, climb, descent, approach and landing, and landing roll segments of the flight. Although it is possible that a momentary power interruption could have caused the solenoid pin to release, which would have then resulted in the pin moving to the extended position and interfering with the ball nut, no conclusive evidence was found during the investigation that supported this possibility. The reason for the failure of the gust lock actuator solenoid could not be determined.

Computed Tomography (CT Scanning) Under the supervision of a lead aerospace engineer assigned to the NTSB's Aviation Engineering Division, located at NTSB Headquarters, Washington, D.C., the rudder gust lock actuator was subjected to x-ray computed tomography (CT) scanning to document its internal condition. The scanning was conducted from December 24, 2014 to March 23, 2015. The scans were performed by Varian Medical Systems, Inc. (formerly Bio-Imaging Research, Inc. (BIR) using a combination of the Varian Actis 500/225 microfocus CT system and the Actis 500/450 standard focus CT system. For the CT scans, the component was loaded into the imaging unit and placed on a turntable. It was then rotated in front of the x-ray source, and the x-rays were captured by a detector after they went through the part. The x-ray source produced a fan beam of x-rays, and the portion of the part imaged was adjusted slightly after each scan was completed until the entire assembly was scanned. The x-ray energy levels captured by the detector were recorded at several thousand different points during each rotation. This information was then converted into slice images using a reconstruction algorithm. The actuator was scanned using a total of 6,421 slices. The total size of the combined data sets was 50.26 Gb. The actuator was scanned multiple times with different scanning protocols using both the microfocus imaging system and the standard focus imaging systems. The microfocus scans provided the best possible spatial resolution, but this type of imaging was constrained to a lower power level that resulted in streaking artifacts within the images. The standard focus scans used a higher power level (with a lower spatial resolution), but these higher power levels eliminated the streak artifacts, and had an inherently higher contrast resolution. Target CT imaging using the microfocus system was used for selected areas to get the highest possible resolution. In addition, the entire part scans and the limit switch scans were re-reconstructed using different reconstruction parameters in an effort to highlight the low density components within the scans. Each data set of slice images was evaluated using the VGStudioMax software package to create orthogonal slice images and a three-dimensional reconstructed image of the component. As part of the evaluation, some sections of the components were digitally removed or rendered transparent to allow closer observation of interior parts. In the images, the high density areas were shown as brighter shades of gray and lower density areas were shown as darker shades of gray. The pointers shown in some of the images denote specific areas of interest within that image. The images of the actuator were examined for any signs of missing or damaged parts, contamination, obstructed passages or any other anomalies. Computed Tomography results revealed the following: • The extend limit switch contacts appeared to be closed during the standard focus scans, and open during the microfocus scans. Whether this difference was due to an actual position change of the contacts or due to the resolution differences of the two scanning methods, could not be determined. • The extend and retract limit switch plungers were extended different lengths (0.87 mm and 1.26 mm respectively). • High and medium density particles were found in various locations throughout the actuator. • A low density indication within one of the extend limit switch screws was only visible in the standard focus scans and not the microfocus scans – this indication is consistent with the apparent low density area being a scanning artifact. • Indications of a crack in the ball screw that were noted in the standard focus scans were more completely resolved in the microfocus scans as the balls and groove located on the ball screw. However, there was an indication of a discontinuity within one of the balls shown on the microfocus images. • There were indications of two areas with distorted windings within the solenoid. Neither of these areas appeared to contain breaks in the windings, nor did either area contact the solenoid plunger • Review of the solenoid images using enhanced contrast methods indicated that there was material containing voids within the passages of the solenoid. • A high density particle was noted within the solenoid within the area of the piston to housing gap. (Refer to the Computed Tomography Specialist's Factual Report, which is appended to the docket for this report.) Rudder gust lock actuator examination by Umbra Cusinetti The rudder gust lock actuator was examined by the manufacturer, Umbra Cusinetti, in the presence of an Italian ASNV representative on June 16, 2015. According to the Umbra Failure Analysis Report, dated November 11, 2015, the results of the examination revealed the following: Initial Visual Inspection • It was observed during the visual inspection that the actuator was locked in a position close to the extended one. Disassembly and Reported Observations • when the actuator was placed on a test rig, it did not move when maximum current at rated voltage was supplied. • when the actuator cover was removed the solenoid/locking system was exposed. • the electrical board was moved to the side but remained connected. • the cam that operates the microswitch was observed closed to the microswitch actuating lever, but did not operate it. • the solenoid was not disassembled. • the solenoid/locking system was observed locked in the extended position. • the Pin position was observed close to the full extended position. Measurement of the exact position was prevented by Pin misalignment due to the distortion of the solenoid plastic reel. • functionality of the microswitch was checked, with positive results. • when the cap was removed it was observed that the ballscrew moved freely. The cap did not show any signs of interference with closed items. • resistance of the three windings were acceptable. • the motor brushes were not checked. • the rotor OD showed signs of possible interference with the stator. • the bearings were found operative with no axial play. Conclusion of the Umbra Cusinetti Examination The solenoid/locking system was found seized at approximately the full extended position Rudder Gust Lock Circuit Board Examination The electronic circuit board within the actuator was examined by the manufacturer, Umbra Cusinetti, on October 14, 2015, in the presence of an Italian ASNV representative. According to the Umbra Failure Analysis Report, dated November 11, 2015, the results of the examination revealed that no evidence of a mechanical failure on components and PCB. Surface coatings have been found damaged. According to the NTSB Systems Group Chairman's Factual Report, Embraer provided the following comments relative to the Umbra Cusinetti examination: Rudder Gust Lock Actuator Position Determination Embraer determined that the rudder gust lock actuator dimensional information gathered during the Umbra Cusinetti examination revealed that with the screwshaft in the "as found" position, the bellcrank would be in contact with the rudder quadrant. It could also engage in the slots on the rudder quadrant. Rudder Gust Lock Actuator Irreversibility, Spring Forces, and Component Replacement - Embraer Evaluation Additionally, Embraer commented that the irreversibility of the gust lock actuator is provided by a plunger that is commanded by a solenoid. In normal operation, the spring is not able to overpower the actuator, since the plunger provides an irreversibility function. The below listed information summarizes the forces required to overpower the actuator motor, assuming the solenoid fails in the energized position which prevents the plunger actuation (resulting in a loss of irreversibility),