ERA23FA235

The pilot’s improper calculation of the airplane’s landing approach speed and required landing distance, which resulted in the airplane touching down fast with inadequate runway available, and a subsequent runway excursion.

The pilot was attempting to land on a wet, 5,000-ft-long asphalt runway in night conditions with calm wind. He stated that the airplane touched down before the 1,000 ft markers and that he immediately applied full brake pressure. The brakes began to cycle in anti-skid mode, but the pilot said the cycling felt slower than normal. The pilot considered a go-around, but the left brake “grabbed” suddenly, and the airplane yawed left and began a sequence of left and right skids before it continued off the end of the runway. The airplane traveled down an embankment onto a rocky berm and came to rest about 360 ft from the end of the runway. The pilot and the passengers egressed the airplane uninjured before a post-impact fire consumed the cockpit, center fuselage, and the right wing. The pilot stated that he used the airplane’s cockpit display unit (CDU) computer to calculate the airplane’s landing reference speed (Vref) of 120 knots (kts) and reported that the airplane touched down at this speed. Although recorded data for the flight was not available due to thermal damage to the avionics, ADS-B data last captured the airplane on final approach at a groundspeed of 120 kts, about 200 ft above the runway, consistent with the pilot’s statement. However, when the conditions that existed at the time of the accident were entered into an exemplar CDU, the calculated Vref was 112 kts, and the required landing distance for a wet/contaminated runway was 4,829 ft. To determine the landing distance required at a Vref of 120 kts, a Vref increment of Vref+5 (117 kts) and Vref+10 (122 kts) were entered into the CDU; the results were 5,311 ft and 5,794 ft, respectively. When these landing distances were calculated, the CDU, which had already been programmed for a 5,000-ft-long runway, displayed a prominent caution below the Vref number that stated: “LANDING FIELD LENGTH INSUFFICIENT.” How the accident pilot achieved a Vref of 120 kts via the CDU based on the conditions that existed at the time of the accident could not be determined. Postaccident examination of the braking system revealed no evidence of any preimpact anomalies or malfunctions that would have precluded normal operation; however, a testing anomaly was observed when the power brake and antiskid valve were tested. An initial test on an exemplar bench stand produced hysteresis that could not be produced when it was tested on a certified bench at the manufacturer. According to the manufacturer, the anomaly observed would not prevent the application of brakes nor the removal of pressure during skidding events; however, significant hysteresis may lead to braking performance degradation due to a decreased pressure application for a given current input. The reason for the anomaly was unknown, and therefore, it could not be determined if it played a role in the loss of braking action reported by the pilot. Despite this anomaly, and based on all other available information, the pilot landed the airplane faster than the prescribed landing speed with insufficient runway length available given the wet runway condition, which resulted in a runway excursion.

HISTORY OF FLIGHT On May 18, 2023, about 0014 eastern daylight time, a Honda Aircraft Company HA-420, N225HJ, was substantially damaged when it was involved in an accident at Summerville Airport (DYB), Summerville, South Carolina. The commercial pilot and five passengers were not injured. The airplane was operated as a Title 14 Code of Federal Regulations Part 91 personal flight. The pilot and passengers were returning to DYB after attending a NASCAR race in Wilkesboro, North Carolina. The pilot stated that he delayed the return to DYB to let rainstorms pass through the Summerville area and knew that as a result he would be landing on a wet runway at DYB. He said that he had landed the airplane at DYB “hundreds” of times on both a dry and wet runway and had never had an issue. The airplane departed Wilkes County Airport (UFK), North Wilkesboro, North Carolina, about 2329, on an instrument flight rules flight plan. The pilot flew the RNAV RWY24 approach coupled to autopilot and took manual control of the airplane around 600 ft above ground level (agl). The pilot said that he calculated the airplane’s approach reference speed (Vref) by entering takeoff and landing data into the airplane’s control display unit (CDU) computer, and that the Vref calculated by the CDU was 119 to 120 kts. He did not recall the required landing distance and said that the CDU gave him an indication that the 5,000-ft runway length at DYB was sufficient. The pilot stated that he flew the final approach at a Vref of 120 kts with full flaps. He stated that he touched down at Vref before the 1,000 ft markers and immediately applied full brake pressure. The brakes immediately began to cycle in anti-skid mode and because of that, very little braking was being applied to slow the plane despite full pedal pressure. The pilot described the anti-skid brakes as cycling slower than normal. The pilot said that it became apparent about halfway down the runway that he would not make the turn off at the second taxiway like he normally did. He said that he was not concerned, because there was plenty of remaining runway and because the airplane had always stopped with runway remaining even when the runway was wet. The pilot added that he considered aborting the landing, but the left brake “grabbed” suddenly, which yawed the nose of the airplane to the left. This began a sequence of left and right skids. The pilot used the rudder pedals to keep the airplane on the runway and was able to straighten the airplane before it continued off the runway into the grass. The airplane slid down an embankment before coming to rest on a rocky berm about 360 ft from the departure end of the runway. The right wing caught on fire, but all six occupants were able to safely exit the airplane through the main cabin door. The airplane was not equipped with a cockpit voice or flight data recorder, nor was it required to be under current Federal Aviation Administration (FAA) regulations. The only recorded data available for the flight was ADS-B data. The data revealed that the airplane’s last recorded position was 200 ft above the runway at 120 kts ground speed and 119 kts calibrated airspeed. PILOT INFORMATION The pilot held a commercial pilot certificate with ratings for airplane single- and multi-engine land, and instrument airplane. His most recent FAA third-class medical certificate was issued on July 22, 2022. He reported about 1,938 total hours of flight experience, of which about 233 hours were in the HA-420. The pilot reported that he last attended recurrent training for the HA-420 in December 2022 at Flight Safety International (FSI), Greensboro, North Carolina. AIRPLANE INFORMATION The airplane was equipped with a Garmin G3000 integrated flight deck, which included multifunction and primary flight displays and an integrated avionics unit. These units and/or their internal data chips were retained for further investigation; however, due to thermal damage, no data were retrieved. METEOROLOGICAL INFORMATION At 0015, the weather reported at DYB included calm wind, 6 statute miles visibility, mist, few clouds at 200 ft above ground level (agl), broken clouds at 3,200 ft agl, overcast clouds at 10,000 ft agl, temperature 21°C, dewpoint 21°C, and a barometric altimeter setting of 29.87 inHg. AIRPORT INFORMATION DYB was a non-tower-controlled, publicly owned airport at an elevation about 55 ft above sea level. The airport had a single runway, 6/24, which was 5,000 ft-long by 75 ft-wide and constructed of asphalt. The runway was described as being in excellent condition. It was crowned, and the asphalt was not grooved. Runway 24 had a downhill gradient about .03%. A postaccident examination of runway 24 revealed witness marks from the accident airplane landing gear. The marks were chalky in appearance and appeared similar to other normal airplane tire marks on the runway. These tire marks were traced from the departure end of runway 24, where the airplane exited the paved surface, back about 1,550 ft, where they ended around taxiway C. The airplane’s touchdown point could not be determined. WRECKAGE INFORMATION The airplane came to rest on a rocky berm on a heading about 240°. A post-impact fire consumed the cockpit, center fuselage, and the right wing. All major components of the airplane were accounted for at the accident site. The outboard section of the right wing had separated from the wing. The majority of the right wing structure was consumed by fire, and the flap setting could not be determined. The No. 2 engine pylon had weakened during the fire and caused the engine to fall toward the airplane. The left wing and No. 1 engine were largely undamaged, except for some impact marks along the leading edge of the wing. The left wing flaps were in the extended/landing position. The aft empennage-mounted speed brake was fully retracted, and minimal thermal or impact damage was observed to the elevator, rudder, and horizontal and vertical stabilizers. The nose landing gear separated from the airplane and was located near the right side empennage. The nose landing gear was largely intact with limited thermal damage. The left main landing gear remained attached to the left wing and was not thermally damaged. There was no rubber reversion or other indications of a skidding tire. The right main landing gear was found in an area significantly affected by fire. The top side of the tire, including the tire sidewall, was thermally damaged. The bottom side of the tire did not appear to be damaged by the heat. The undamaged part of the tire revealed no rubber reversion or other indications of a skidding tire. Postaccident examination and functional testing of both brake assemblies, three of the four brake master cylinders, the brake shutoff valve, both wheel-speed transducers, and the emergency parking brake valve revealed no anomalies. The antiskid control unit and one of the brake master cylinders were thermally damaged and could not be functionally tested. Initial functional testing of the power brake and antiskid valve was performed at HondaJet and revealed an unexplained hysteresis when plotting the antiskid current input against the left brake pressure output. The unit was subsequently sent to the original equipment manufacturer for functional testing on their established test equipment. The unit passed all portions of the functional test, including the hysteresis test. A computed tomography scan and subsequent disassembly of the unit revealed no indications of internal damage or foreign object debris. It was not determined if the initial test setup or equipment used at Honda contributed to the initial behavior, and a root cause for the initial behavior could not be determined. According to the brake manufacturer, the observed initial behavior would not prevent the application of brakes nor the removal of pressure during skidding events; however, significant hysteresis