DCA22LA182

A lateral runway excursion during landing for reasons that could not be determined based on the available evidence.

After a normal touchdown and initial deceleration, the airplane tracked the runway heading before beginning a right turn to exit on a high-speed taxiway. The flight data recorder (FDR) revealed that, as the right turn developed, the captain released the brakes and the cockpit voice recorder (CVR) revealed that the captain commented about the lack of nosewheel steering. The captain applied full left rudder and used the nosewheel steering tiller but was unable to arrest the right turn, which continued but at a slightly reduced rate, until the airplane departed the paved surface and came to a full stop in the grass to the right of taxiway NK. During the excursion the airplane struck an airport sign and sustained minor damage to the right wing flap skin and fairing. Neither the tiller control input nor the nosewheel steering (NWS) angle were recorded by the FDR. The thrust reversers remained deployed throughout the turn with a slight asymmetry. Although this would have provided some right-turning tendency, lateral control can be maintained with one engine at idle and the other at maximum reverse thrust using rudder and differential braking. The wind gusts recorded at the airport at the time of the incident could have produced a right crosswind component up to 15 knots. Given the airplane’s crosswind component limitation of 27 knots, the slight thrust asymmetry or crosswind component would not have rendered the airplane uncontrollable. Post-incident testing of NWS components revealed no anomalies that would have precluded normal operation of the system. However, the handwheel control unit (tiller) failed the initial return-to center acceptance test. It was then cleaned, retested and passed. Although the FDR recorded a Nose Wheel Steer Status “FAIL/Not Armed” condition, and the crew observed a “STEERING INOP” message on the engine indicating and crew alerting system display, both of those messages occurred after the airplane departed the runway. The cause of these indications could not be determined. One possible condition that would have caused these indications was rapid tiller inputs, but the investigation could not assess those inputs because they were not recorded on the FDR. Several fault codes were recovered from the NWS electronic control unit (ECU). Although these codes have no date or timestamp, they are stored chronologically. The most recent code recorded was “PILOT_HANDWHEEL_RATE_FAULT,” which, according to the type certificate holder’s Fault Isolation Manual, can occur if (1) the handwheel (tiller) is moved too rapidly or (2) the handwheel is moved (or was in a position) too far from center immediately after the extension of the landing gear. While it is likely that at least one code would have been logged on the ECU due to the recorded “FAIL/Not Armed” condition captured by the FDR, had this been the only code associated with the event, it alone, nor its potential causes, can explain the loss of lateral control. The FDR recorded “FAIL/Not Armed” condition occurred after the airplane departed the runway and may have been caused by the captain’s tiller inputs, or erratic nosewheel movement due to the excursion. Neither of which could be confirmed or refuted. Given that the airplane initially tracked along the runway heading immediately after landing, it is unlikely that the nose landing gear steering angle would have been off center before or upon touchdown. The pilot used some differential braking to counter the right turn, though not to the fullest extent available. The pilot’s left rudder applications, which would have been less effective as the speed decreased, should have also commanded the nosewheel to turn up to 7° to the left of the airplane’s centerline. The pilot reported that he also used the tiller in an attempt to arrest the right turn, though a lack of recorded data precluded a determination of when and how much tiller command was applied. Post accident maintenance evaluation of the nosewheel steering system by the operator and subsequent component testing at manufacturers’ facilities found no evidence of malfunctioning or inoperative components. The cause of the right turn and why the pilot’s inputs were insufficient to arrest it, could not be determined.

According to the Mitsubishi Heavy Industries Regional Jet – (MHIRJ, the type certificate holder) Flight Crew Operating Manual, the nosewheel can be steered up to 70° left or right via the tiller handwheel and up to 7° left or right via the rudder pedals. Further, if a failure is detected by the nosewheel steering (NWS) electronic control unit (ECU), which is also known as the analog/digital steering controller, or if hydraulic system No. 3 fails, the NWS would revert to a shimmy-damping mode to allow free-castering of the nosewheel. The pilot could then maintain directional control on the ground using the rudder, differential braking and differential thrust. The ECU captures detected fault codes in nonvolatile memory (NVM). The codes are stored in reverse chronological order without dates or timestamps. According to the MHIRJ Fault Isolation Manual, NWS is deactivated when the weight on wheels (WOW) signal indicates “air.” When the airplane is airborne, the steering command signals are replaced with a “straight ahead” reference. Also, a mechanical cam mechanism inside the nose landing gear (NLG) strut ensures that the NLG is straight when the strut is fully extended. FDR data indicated that the WOW switches and hydraulic system No. 3 pressure were operating normally. If the tiller handwheel is not centered during touchdown – and the nosewheel steering angle is less than 7°, the ECU would not detect a fault. If this situation were to occur, when the nosewheel WOW switch changes to ground, the ECU would combine the tiller commanded position with the rudder pedal position, to compute a steering command for the nosewheel. The NWS Status FDR parameter changes to FAIL/Not Armed if the ECU detects a fault with one or more NWS system components or the NWS is not armed. This condition is annunciated on the EICAS display in the cockpit as a caution (amber) “STEERING INOP” message with an accompanying aural chime. Faults include system component failures; improper voltages; degraded performance; or unexpected operating conditions such as valve positions inconsistent with commanded position, handwheel commanded steering angle (greater that 7°) during touchdown, or excessive rates of operation, that is, moving the tiller handwheel or rudder pedals too quickly (sensors rotated more than 500°per second). According to the SkyWest Dispatch Procedures Manual, the airplane’s crosswind component limitation was 27 knots for a dry runway. Post-incident Examinations After the event, SkyWest maintenance downloaded the NVM from the ECU, which contained the NWS fault codes shown in table 1.Because no dates or times were associated with these codes, the investigation could not directly determine which (if any) of the codes were related to the event. However, the codes are stored in the sequence in which they occurred in (reverse) chronological order. The table below was provided by MHIRJ, the potential cause(s) of the codes are referenced from the MHIRJ Fault Isolation Manual STEERING INOP Message section. The most recent code is listed first. Code Description Potential cause 033 PILOT_HANDWHEEL_RATE_FAULT Detection of handwheel position greater than 8° from center during a 12.5-second system test that occurs immediately after nosewheel extension. Movement of the handwheel at a rate greater than 500° per second 191 WARM_BOOT_OCCURRED ECU 115 AIR_TO_GROUND_MODE_CMD_FAULT Handwheel- commanded steering angle (upon air-to-ground transition) exceeds 7° of NWS angle 109 SSV_STUCK_CLOSED_PS1 SSV (solenoid selector valve), or PS1 (pressure switch) or NLG (nose landing gear) priority valve or NLG selector valve or NWS system check valve blockage or wiring 170 PILOT_HANDWHEEL_RVDT_FAULT Pilot’s handwheel control unit or wiring 036 FEEDBACK_TRACKING_FAULT R feedback RVDT (rotary variable differential transformer) or L feedback RVDT or grease contamination or steering collar wear 105 SSV_COIL_BIT_FAULT SSV or ECU or wiring or NWS system check valve blockage Table 1: Nose wheel steering electronic control unit fault codes. (Source: MHIRJ) SkyWest maintenance also performed the instructions in the MHIRJ Fault Isolation Manual for the STEERING INOP caution message but was unable to determine why the message occurred. A timing check was performed to determine whether the NWS handwheel would return to center from the full-left and full-right positions within the 2- to 4-second range, and the check showed that recentering was within the specified range. SkyWest maintenance then replaced the NWS handwheel control unit (tiller) and ECU, checked the voltages for the left and right actuator feedback rotary variable differential transducers (RVDT), and the rudder pedal RVDT; all were within limits, and the NWS system operated normally (RVDTs are angular position sensors). The ECU, as well as the handwheel control unit, electrohydraulic servo valve, steering selector valve, rudder RVDT, a wiring harness, hydraulic manifold, and compensator were shipped to their respective manufacturers/suppliers for examination and acceptance test procedure (ATP) testing. The ECU, rudder RVDT, and compensator passed all ATP test elements. The handwheel control unit was tested at its supplier’s (Safran Landing Systems) facility before the start of the NTSB’s investigation, and no test report was generated. Safran reported that the unit failed the return-to-center element of the ATP. The unit exceeded the allowable 4 seconds to return to center after it was fully displaced in both directions (left and right). The unit was cleaned, and then it passed the ATP test. The electrohydraulic servo valve operated normally and met ATP specifications except for null bias. The null bias shift was small (measured at 0.21 mA, maximum allowable was 0.16 mA). The steering selector valve passed all functional elements of the ATP but failed one nonfunctional element (the insulation resistance test at 500 VDC). The resistance criteria was 500 megohms or greater; the tested value was 7 megohms. The NLG wire harness passed all ATP elements except for the dielectric strength test at 550 VDC. Disassembly of the harness found two broken shield wires, which are used for a health monitoring function and do not affect steering. The hydraulic manifold passed all ATP elements except for the “Flow1” test, which required the hydraulic fluid flow to be between 0.95 and 1.40 gallons per minute. The flow was measured as 1.41 gallons per minute. None of the minor discrepancies would have prevented normal operation of the NWS system. Maintenance Records A review of the airplane’s maintenance records over the 12 months preceding the incident related to the NWS system revealed the following pilot-reported discrepancies: One report of the tiller not being centered and the airplane pulling “excessively” to the right during taxi, along with the STEERING INOP EICAS message. The discrepancy was cleared after the handwheel control unit was inspected, and the nose wheel steering health check was performed. Neither of those tasks found any anomalies. Six additional reports of the STEERING INOP EICAS message were noted. Three were cleared after performing a reset of the NWS system, two were cleared after NWS system operational checks were performed, all with no anomalies found. One was cleared after the right feedback actuator RVDT was found out of rig. A review of the maintenance records related to the NWS over the 6 months following the incident revealed a pilot-reported discrepancy on December 15, 2022, which stated, “excessive pull to the left and over correction with both tiller and rudder.” After the report, the left feedback actuator RVDT voltage was found to be out of tolerance. The discrepancy was cleared after replacement of the RVDT. A second pilot-discrepancy occurred on December 19, 2022. The discrepancy stated, “on ta