Mesa, AZ, USA
N150NE
ECLIPSE AEROSPACE INC EA500
The accident was the first flight since the airplane had undergone maintenance, which was completed earlier in the day. After an uneventful flight, the pilots began configuring the airplane for landing on the runway, which had a 0.6% downslope, and according to automated weather reports from the airport a 4-knot tailwind. The pilot-in-command reported that the airplane touched down in the landing zone, just beyond the 1,000-ft markings, but that when he applied brake pressure, the airplane did not decelerate and started to drift to the left. The co-pilot then began to also apply brake pressure, and with no significant deceleration and the runway end approaching, he called for the pilot to activate the “ALL INTERRUPT” switch, to which the pilot responded that he already had. The pilot stated that he was also applying full right pedal to try to steer the airplane back toward the runway centerline, but the airplane did not respond. It eventually departed the left side of the runway, crossed over the last two taxiways, and came to rest in a drainage culvert, resulting in substantial damage. Review of ADS-B and recorded onboard data showed that, although the approach path was relatively stable, the airplane’s landing reference and touchdown speeds were about 20 knots above that recommended for its configuration and the runway conditions. Additionally, the data indicated that, at the time of touchdown, the tailwind had likely increased to about 10 knots. Recorded acceleration data indicated that the airplane touched down about 1,300 ft from the runway threshold, consistent with the pilot’s statement; however, due to lift still being created by the airplane’s wings because of the excessive speed, braking action would have been limited. This was corroborated by the weight on wheels (WOW) sensors, which did not activate until the airplane had progressed 2,490 ft down the runway. At this point, the airplane would have had about 2,760 ft of the runway surface remaining, including the blast pad, along with a 500-ft level dirt runoff area. Although performance calculations indicated this would not have been enough runway surface to stop the airplane, extrapolations based on the deceleration data attained from multiple previous flights showed that, with operational brakes, the airplane may still have been able to stop on the 150-ft blast pad, assuming it had not departed the runway to the side. Examination of the runway surface revealed indications of predominantly left tire marks starting just left of the runway centerline about 3,300 ft beyond the landing threshold. The marks continued for the remaining length of the runway, gradually veering left until crossing over a taxiway and becoming darker as they approached the drainage culvert. Examination of the hydraulic brake system did not reveal any evidence of pre-accident failure; however, the right anti-skid brake system (ABS) wheel speed transducer (WST) showed evidence that it had not been properly aligned with the hub cap when the cap was installed following maintenance. Maintenance instructions stated that such a misalignment can damage the brake system and interfere with braking, and subsequent testing revealed that the WST was damaged and not functioning correctly. Under this condition, the ABS would have assumed the right wheel was locked in a skid and released brake pressure, rendering the right brake inoperative. This would have resulted in the degraded brake performance reported by the pilot, and the airplane’s transition to the left side of the runway during the landing roll as observed. The airplane’s emergency procedures stated that, should the brakes become ineffective, or the airplane pulls to one side, the pilot should press and hold the ALL INTERRUPT switch until braking action was restored. Activation of this switch with a failed WST would have disabled the ABS system and restored normal braking. Although the pilot stated that he used the switch as the airplane approached the end of the runway, the onboard data appeared to indicate that the last time it was used was during the approach, when the pilot briefly activated it to confirm that the yaw damper was off.
HISTORY OF FLIGHTOn November 3, 2022, about 1915 mountain standard time, an Eclipse Aerospace EA500, N150NE, was substantially damaged when it was involved in an accident in Mesa, Arizona. The pilot and co-pilot were not injured. The airplane was operated as a Title 14 Code of Federal Regulations part 91 post-maintenance positioning flight. The airplane underwent maintenance at a facility at Henderson Executive Airport (HND), Henderson, Nevada. The maintenance was completed on the day of the accident, and the accident flight was the first flight after the airplane was returned to service. The work performed included the completion of the 24-month/300-hour airframe inspection. Both pilots reported that the brakes appeared to operate during preflight and taxi operations, although due to rain and the airplane not being equipped with windshield wipers, they were traveling relatively slowly during the taxi. The pilot reported that, after an uneventful flight, he and the co-pilot began configuring the airplane for the landing on runway 22L at Falcon Field Airport (KFFZ), Mesa, Arizona. After slowing the airplane and configuring the flaps and landing gear, they were cleared to land. He reported that, as part of the before-landing checklist, he confirmed that the ABS was armed and pressed the ALL INTERRUPT switch to ensure the yaw damper was off. He reported that the airplane touched down in the landing zone just beyond the 1,000 ft markings, but when he applied brake pressure, the airplane did not decelerate and started to drift to the left. The co-pilot then began to also apply brake pressure; however, with no significant deceleration and the runway end approaching, he called for the pilot to activate the ALL INTERRUPT switch, to which the pilot responded that he already had. The pilot stated that he was also applying full right pedal to try to steer the airplane back towards the runway centerline, but the airplane did not respond. It eventually departed the left side of the runway, crossed over the last two taxiways, and came to rest in a drainage culvert. METEOROLOGICAL INFORMATIONAn automated observation for FFZ was issued at 1854 and indicated clear skies, visibility 10 statute miles, wind from 040° at 4 knots, temperature 11°C, dewpoint 3°C, and an altimeter setting of 30.04 inHg. The reports at 1910 and 1915 showed that the temperature, dewpoint, and altimeter setting were unchanged. At 1910, the reported wind was from 050° at 5 kts and, at 1915 (the time of the accident), wind was from 050° at 4 kts, and nearly a direct tailwind for runway 22L. WRECKAGE AND IMPACT INFORMATIONRunway 22L was examined for evidence of tire transfer marks. The first identified mark was from the left tire, just left of the runway centerline about 3,300 ft beyond the landing threshold. The mark continued for the remaining length of the runway, gradually veering left until it crossed over the side stripe at taxiway D2. Three distinct wheel marks were then evident as they crossed over a grass area and through taxiway D1. As the marks continued across the taxiway, the left tire track became distinctly darker, continuing beyond the taxiway and into a drainage culvert at the end of the runway. The airplane came to rest in a 6-ft-deep drainage culvert 115 ft beyond the end of runway 22L and about 180 ft left of the runway centerline. The airplane sustained damage to the nose structure and both flaps. The nose and right main landing gear assemblies had partially collapsed into their respective wheel wells. Both the nose wheel and the left wheel and brake assembly had broken away from their respective trunnions, and the nose tire had separated from its rim. There was no evidence of heat discoloration or fluid/oil contamination to the brake pads or disks, all of which displayed normal wear and had friction surface life remaining. There was no evidence of pre-accident brake line damage. The right brake was functionally tested with brake pedal application and was operational; impact damage prevented a similar assessment of the left brake system. The left tire had worn down to between 1/32nd and 1/8th inch tread depth; the right tire appeared relatively new with ¼ inch tread depth remaining. Both tire contact surfaces and side walls exhibited diagonal and longitudinal scratches and scuffs, and no bald spots were evident. The left and right tire pressures were 82 and 86 psi, respectively. ADDITIONAL INFORMATIONThe Emergency section of the airplane’s Airplane Flight Manual (AFM) provided procedures when the brakes are ineffective, or the airplane pulls to one side, specifically: 1. Maintain directional control using rudder and steering 2. Brakes: Release 3. ALL INTERRUPT: PRESS and HOLD 4. Reapply Brakes (Pump Brakes as required): Stop Normally NOTE: ABS Functionality Unavailable The Description and Operation section of the AFM stated: If emergency disablement of the ABS is needed, the pilot (or co-pilot) may press and hold the ALL INTERRUPT switch which will command the ABS to discontinue its function and restore normal braking. If the ALL INTERRUPT switch is held during heavy braking, the ABS will be inoperative and tires may skid until brake pressure is reduced by the pilot. When the ALL INTERRUPT switch is released the previous ABS state will be restored. Performance Data An aircraft performance study was performed by a specialist in the NTSB’s Vehicle Performance Division. Review of the DSU data indicated that the airplane departed HND about 1823. The flight progressed on a southeast track, reaching 23,000 ft about 10 minutes after takeoff. At 1902, the pilots set the altimeter to 30.04 inHg, consistent with the FFZ ASOS report. The flaps were set to 15° degrees at 1908:35, and the landing gear extended at 1910:42, followed 30 seconds later by the flaps being set to 32°. The airplane track aligned with the runway at 1913:00, about 2 minutes before touchdown. At that time, the airplane was about 3.4 nautical miles from the threshold of runway 22L, and about 1,325 ft above the field elevation. The airplane was initially below the 4° glidepath defined by the PAPI lights, but from about 1914:30 and 4,500 ft from the threshold, the descent became aligned with the glidepath. The descent rate for the final 2 minutes varied between about 400 and 800 fpm and was about 600 fpm as the airplane crossed the runway threshold. According to the EA550 landing distance chart, using the airplane’s weight and landing flap configuration with the runway parameters and a 4-kt tailwind, the expected landing distance would have been about 3,370 ft. The airplane’s landing reference speed (Vref) should have been about 87 knots equivalent airspeed (KEAS), and touchdown speed (Vtd) 75 KEAS. According to the DSU data, at the threshold, the equivalent airspeed was about 18 kts above Vref. Touchdown was identified by spikes and increased noise in the recorded acceleration at 1915, and at that time, the equivalent airspeed was about 20 kts above Vtd. The data recorded at that time showed the groundspeed as about 106 kts, indicating a tailwind of about 10 kts at touchdown. Based on acceleration data, touchdown occurred about 1,300 ft from the threshold, leaving 3,800 ft remaining on the runway, plus 650 ft for the blast pad and overrun area. The DSU data indicated that the left WOW sensor transitioned to ground mode about 2,490 ft from the threshold, by which time the EAS was 80 kts (groundspeed 88 kts). The data corroborated the pilots’ account of the airplane veering to the left of the runway centerline as the runway end approached and suggested that the airplane was still traveling about 40 kts as it veered left and departed the runway surface. The DSU data showed that the ALL INTERRUPT switch was toggled during the approach phase at 1910:25, and again at 1913:30 consistent with the pilot’s statement; however, there was no recorded discrete event or data indicating that the switch was operated at any time during the landing roll. Review of the last 10 flights before the accident showed that the airplane routinely exceeded Vref by speeds averaging about 10 kts and ranging between 6 and 16 kts. Irrespective of the airplane’s performance charts, extrapolations based on the deceleration data attained from the ten previous flights show that with operational brakes, the airplane may still have been able to stop on the 150-ft blast pad beyond the runway, assuming it had not departed the runway to the side. Various Federal Aviation Administration sources provide general guidance on stabilized approaches. This guidance suggests that for an approach under visual flight rules, a stabilized approach includes having the airplane configured for landing and maintaining a consistent and controllable descent at or above 500 feet above the touchdown zone while on the final approach course. Maintaining the correct flight path should require only small corrections, the aircraft should be within +10 to -5 kt of the recommended airspeed, and it should be descending at less than 1000 ft/min. FLIGHT RECORDERSThe airplane was equipped with an integrated diagnostic storage unit (DSU) capable of recording GPS position along with engine and airframe parameters at various intervals. Neither control surface positions (except flaps), nor parameters regarding brake application or brake pressure were recorded. The airplane was not equipped (nor was it required to be equipped) with a flight data recorder or cockpit voice recorder. TESTS AND RESEARCHBrake System Operation The braking system was conventional and consisted of brake master cylinders coupled to the toe brakes and connected via hydraulic lines to brake calipers on each main landing gear. Each caliper contained five hydraulic pistons, which pressed pads against a single disk. There are no brake boosters or other pressure assist systems, and the airplane was not equipped with thrust reversers or spoilers for use in landing. The ABS system consisted of a brake controller (DECU), a left and right brake control module (BCM), and WST fitted to each the wheel axle. Although the airplane was equipped with WOW sensors, they were not interfaced with or used by the braking system. Aircraft ground speed (derived from GPS data) and wheel speed are used by the DECU to determine the difference between wheel speed and aircraft speed. If the DECU senses a skid, it commands the BCM to release brake pressure until the wheel speed matches the GPS speed. Both BCM’s operate independently of each other. Each BCM contains a series of hydraulic valves and a motorized pump. The modules are fitted in the feed line from the master cylinder to the calipers on each side. The WST within the wheel axle has a keyed tab, which engages with a slotted drive adapter mounted inside the hub cap. The ABS system does not activate during normal ground and flight operations and is disabled during low-speed taxi. When the brakes are applied and ground speed is above 15 kts, the BCM activates and provides pressure (charge) to the pressure side of the module through activation of the motorized pump. This motor operation may be heard in the cabin, but does not indicate that the ABS is active. Once the brakes are applied during landing or taxi and a skid is detected, brake pressure is directed back to the master cylinders and released from the brake caliper to the wheel, allowing the tire to regain its speed. The BCM will stop pushing against the master cylinder(s) and brake pressure will be restored once the wheel is accelerated above the skid threshold by runway friction. The cycle will repeat for as long as the pilot applies brake pressure beyond that required to skid the tire. The airplane was equipped with a model of WST that used an adhesive to bond its input shaft to the drive bore. This type of transducer was the subject of Eclipse Aerospace Service Bulletin (SB) 550-32-005, Rev B, which was intended to ensure that the drive shaft was adequately bonded to the bore and could not be pulled out. The bulletin provided instructions for inspection, test, and repair (if necessary) of the transducer. There was no reference in either the airplane’s maintenance logbooks or Eclipse Aerospace service records to indicate the service bulletin had been complied with on the accident airplane. The airplane was used in both Part 91 and Part 135 operations, and compliance with the service bulletin was not mandatory. Later revisions of the WST used a keyed rather than bonded shaft. Damage to the landing gear prevented an accurate assessment of the WOW switch rigging and operation. An ABS indicator is included on the left side of the instrument panel. It provides status of the system (ARMED or INOP). The INOP indicator illuminates when any of the following faults are detected: GPS ground speed is out of range (< 0kts or >500kts) or unavailable, wheel speed is out of range (>200kts) or unavailable, the controller power is low, the controller Built-in-Test (BIT) has failed, or the ABS BCM solenoid control line is in open circuit. There are no annunciators that signal if the ABS is actively operating during a skid condition. Both pilots reported that there had been multiple occasions in the past when they felt ABS activation during braking, and that at no point during the event did they feel such activation. The before-landing checklist included a warning that the brakes should not be applied in the air with the ABS system armed, otherwise a loss of braking pressure would occur that could only be overcome on the ground with pumping application of the brakes. However, this warning was not applicable to the accident airplane, which included an upgraded BCM. All Interrupt Switch The ALL INTERRUPT switch is used to interrupt any non-pilot-activated input to the flight control system. It consists of a red button mounted to each sidestick. The switches disconnect the autopilot, yaw damper & flight director, and interrupts the ABS, stick pusher and all trim while held. Should a switch fail, an advisory is displayed. The pilot stated that he was certain that he pressed the ALL INTERRUPT switch just before the airplane departed the runway. Brake System Examination The right WST was examined through the access panel on the hub cap using a lighted borescope. The sensor drive spline was engaged with its drive slot adapter in the hub cap. The hub cap had sustained light scratches to its outer surface but was otherwise undamaged. Removal revealed that the adapter was intact and free of damage, but there was evidence of material transfer and “smear” marks between the WST drive spline and the drive slot of the adapter. The material smear was opposite the direction of hub rotation (see figure 1). Further examination revealed matching transfer mark damage to the drive spline and that it was loose and could be pulled out of from the sensor case. Figure 1 - WST spline and hub cap adapter. Primary ABS components, including the WSTs, brake control modules, DECU, and low pressure switches were examined at the system’s manufacturing facility under the oversight of the NTSB. All examined components, except for the right WST, either met the manufacturer’s acceptance test standards or showed results consistent with normal in-service use. During the testing of the right WST, although the splined end of its input shaft moved freely when rotated by hand, it could be easily pulled out of the assembly (see figure 2). As a result, although the WST passed acceptance tests when the shaft was in place, it stopped producing a valid signal once it was pulled out by about 1/8 inch. Figure 2 – Splined shaft detached from WST assembly. The airframe inspection that was completed on the day of the accident would have necessitated the removal of the wheels and hub caps. The reassembly instructions included in the applicable maintenance manual provided specific instructions for alignment of the WST keyed tab with the slotted adapter in the hub cap, along with the following warning: CAUTION: MAKE SURE TO ALIGN THE BRAKE ASSEMBLY TABS WITH THE SLOTS ON THE WHEEL ASSEMBLY. IF THE BRAKE AND WHEEL ASSEMBLY ARE NOT KEYED TOGETHER
Maintenance personnel’s failure to properly align the right brake wheel speed transducer, which rendered the right wheel brake ineffective during landing, and the pilot’s subsequent failure to follow the correct procedures to restore braking action during the landing roll. Contributing to the accident was the pilot’s failure to attain the proper approach and landing speed, which resulted in the airplane landing excessively fast.
Source: NTSB Aviation Accident Database
Aviation Accidents App
In-Depth Access to Aviation Accident Reports
