Sierra Vista, AZ, USA
CBP113
GENERAL ATOMICS MQ-9
At the time of the accident, the flight was operating as a training evaluation mission for a Launch-Recovery Element (LRE) checkride, performing a sequence of touch and go landings on runway 26. The accident landing was the first during this session using the Multi-Spectral Targeting Ball (MTS-B) payload camera vice the fixed nose camera. The payload camera, located about 3 feet lower than the fixed nose camera, produces a different angle and field of view. Additionally, the payload camera is gyro stabilized for mission requirements and gives a different visual perception during landing. The flare during the accident landing was initiated closer to the runway and to less of a nose up attitude than normal. The aircraft bounced and contacted the runway four times, the magnitude of the bounces and pitch excursions dynamically increasing until the nosewheel fractured and the aircraft began to slide. During the bounces, the pilot began to initiate a go-around and at the same time, the evaluator pilot took hold of the control stick in an attempt to correct the bounces; however, the corrective action was too late to prevent the accident. There was no evidence of any pre-existing damage to the nose gear components and no anomalies in the flight controls.
HISTORY OF FLIGHT On November 6, at 2236 mountain standard time, a General Atomics Predator-B MQ-9 unmanned aircraft, registration CBP113, controlled from Ground Control Station GCS 3001, departed the runway after a hard landing at the Libby Army Airfield, Fort Huachuca, Sierra Vista, Arizona. At the time of the accident, the flight was operating as a training evaluation mission for a Launch-Recovery Element (LRE) checkride. Prior to the training exercise, the aircraft had been performing a Customs and Border Protection (CBP) operational mission and control was handed over to the training pilot and evaluator. The aircraft was performing a sequence of touch and go landings on runway 26. The first landing attempt was too high and the pilot executed a 360 degree turn to reposition for landing, then proceeded to a planned go-around. The first two landings were conducted using the fixed nose camera. The pilot and evaluator pilot (EP) discussed the first landing during the subsequent traffic pattern, and the second landing was much smoother. On the third landing of the sequence, the aircraft touched down on the nosegear and porpoised. The accident landing was the first during this session using the Multi-Spectral Targeting Ball (MTS-B) payload camera. During the downwind leg of the third pattern, the EP instructed the pilot to switch the view to the payload camera. The aircraft appeared stable on short final, the pilot was maintaining a reference speed of 90 knots, which slowly reduced to about 75 knots just prior to touchdown on the nosegear. Recorded data indicates the flare was initiated closer to the runway and to less of a nose up attitude than on the previous landing. The aircraft bounced and contacted the runway four times, the magnitude of the bounces and pitch excursions dynamically increasing until the nosewheel fractured and the aircraft began to slide. The first impact was at a one degree nose low attitude and resulted in a small increase in pitch. The pilot then input forward stick and the aircraft touched down again, with slightly more force and pitched up to about 5 degrees in response. The pilot input more forward stick, repeating the cycle, and following the third bounce, the recorded data indicates the control stick moving aft to about 5 degrees. However, at that point the aircraft pitch attitude had begun to trend downward and the aircraft impacted as much as 15 degrees nose down. Another pitch cycle ensued. However, by this point the nosewheel had likely fractured, and the following impact damaged the payload ball. The pilot reported that he began to initiate a go-around and at the same time, the EP reported that he put his hand on the control stick to ensure that it was in the correct position. He then pulled it aft in an unsuccessful attempt to correct the bounce. The EP noted that it was possible to disengage the control stick with pressure from above on the trim button, but could not tell if that occurred during the accident landing. The EP noted that he observed the pilot had a low pitch attitude, and that due to a number of previous tailstrike events with the MQ-9 fleet-wide, pilots tend to not want to flare very much. Airport fire crews responded to the aircraft following a call from the ATC Tower responding to the sensor operator, and two General Atomics technicians responded to the aircraft to shut off on-board batteries and engine controls. DAMAGE TO AIRCRAFT The nosewheel rim fractured, and nose tire burst. The wheel assembly separated from the nose strut. The left main gear was bent approximately halfway along the main strut and the right main gear failed upward and outward. The MTS-B was abraded and damaged. The lower vertical stabilizer separated from the airframe. The propeller was bent and about 10 inches of one blade separated from the propeller. OTHER DAMAGE There were some small gouges in the runway from the fractured propeller blade stub. PERSONNEL INFORMATION The pilot was receiving a checkride for qualification for LRE operations. He had been recommended for the checkride the previous evening after completing “Lesson 15 Checkride Prep” in the LRE syllabus. Comments were favorable, and noted “excellent job doing the night nose camera landing.” He had practiced landings using both the nose camera and the MTS-B infrared camera within the preceding few lessons. He was a certified pilot in manned aircraft, with approximately 9000 hours total time, and flew the Citation and other aircraft for CBP. He was new to the UAS and transitioned to the Predator in March of 2008. He was a qualified Mission Control Element (MCE) pilot, with about 100 hours in the MQ-9. The Evaluator Pilot (EP) was a contractor to General Atomics. He had previously worked for GA flying various UASs, and left to form his contracting company. He provided training and pilot support to the CBP. He had over 5400 hours PIC on UASs and over 5000 landings, and was initially certified in April of 1997. He had logged over 1600 hours on the MQ-9. He was an FAA certified commercial pilot in manned aircraft and flew various types of general aviation aircraft. The sensor operator was fully qualified in his position, and had worked for GA for 2.5 years. He was a former US Army Signals Intelligence specialist before being employed by GA. He was certified on both the MQ-1 and MQ-9, with about 1800 total hours and 460 hours on the MQ-9, and had worked in Iraq and the US. The sensor operators support the UAS pilots in executing checklists, making callouts, etc. as determined by the crew in a pre-flight briefing. The Command Duty Officer was an MCE rated MQ-9 pilot. The CDO function serves as an on-site government representative, conducts crew briefing, coordination with weather and ATC and other coordination duties. The CDO on duty at the time of the accident had been with CBP for over 9 years, and was a certified helicopter pilot also. AIRCRAFT INFORMATION CBP113 was built in 2007 and delivered to CBP in 2008, it was one of four Predator-B MQ-9 UAVs operated by CBP. CBP records indicate the aircraft had approximately 444 flight hours prior to the accident flight. The GCS was one of two located at Libby Army Airfield, two others were located at Riverside CA, and Grand Forks ND. CBP113 was configured for airborne reconnaissance in support of CBP missions. The MQ-9 was not certified by the FAA, but operates under the provisions of a Certificate of Authorization issued by the FAA. Although the aircraft is not certified, General Atomics stated that they meet the provisions of 14 CFR Part 23 whenever practicable. The MQ-9 was powered by a Honeywell TPE-331 turboprop engine. The MTS-B payload camera was located on the underside of the nose of the aircraft, about 6 inches off centerline to the right. It was about 3 feet lower than the nose camera, which was fixed in the forward fuselage aligned to the centerline of the aircraft. The payload camera had a slightly narrower field of view, and was a higher resolution picture. The payload camera was gyro stabilized for mission requirements, and pilots reported that it gave a different perception upon landing touchdown. The aircraft was described by pilots as very “glider-like” in performance during the landing maneuver. The MQ-9 trim system was not actuated by a conventional “coolie hat” or wheel. When a pilot wished to trim the aircraft for hands off flight in a certain attitude, he positioned the control stick to achieve the desired flight attitude, then depressed and held a switch on the top of the stick. After returning the stick to the center detent, releasing the switch recentered the system to be in trim at the pilot-selected attitude. Whenever the switch was depressed, movement of the stick would not command flight control surface changes. Takeoff weight of the aircraft was 7,827 pounds, planned landing weight was 6,400 pounds. METEOROLOGICAL INFORMATION Weather the time of the accident was reported as light and variable wind, sky clear, visibility greater than 10 miles, temperature 19 degrees C, dewpoint -10 degrees C, sea level barometric pressure 30.14”. AIDS TO NAVIGATION The aircraft navigated via GPS. There were no anomalies evident, no receiver autonomous integrity monitoring (RAIM) warnings, and no satellite outages. There were no anomalies noted in data uplink/downlink between the GCS and UA. COMMUNICATIONS Communications with ATC were nominal. AERODROME INFORMATION Libby Army Airfield is located within the city of Sierra Vista, Arizona. It is a joint use civil/military field. The runway in use, 8/26, was 12,001 feet long, 150 feet wide with full safety areas on each end. Runway 26 touchdown zone elevation was 4,599 feet, and the runway sloped at a 1% uphill grade toward the west. The runway condition was dry at the time of the accident, there were no reported NOTAMs of significance. The Air Traffic Control Tower was in operation at the time of the event, and there were no other aircraft operating in the traffic area as required by the CBP Certificate of Authorization. No abnormalities regarding the airport or ATC were reported. FLIGHT RECORDERS The MQ-9 GCS was equipped with a full data logger, which recorded approximately 1900 parameters of command and status information from the aircraft and ground station. Additionally, video recordings from both the fixed nose camera and MTS-B payload camera were recorded. All data was downloaded successfully. The UAS does not record any audio information. WRECKAGE AND IMPACT INFORMATION The first piece of nosewheel rim was located on the runway at about 2300 feet from the approach threshold, 20 feet to the left of centerline. Numerous other pieces of nosewheel rim were found along the runway for about 100 feet. A large mark consistent with nosewheel rubber and rim metallic impact was found at the beginning of a long scrape mark along the runway slightly right of centerline. Five prop strike marks were found in the same area. Debris and runway scarring were consistent with the lower vertical stabilizer contacting the runway, shedding some small pieces of fairing and subsequently separating from the airframe. Just beyond the position of the separated vertical stabilizer, markings indicate the aircraft began to swerve left, and exited the runway. Ground scars in the grass were consistent with the leftward rotation of the airframe to the final orientation. The payload ball was severely damaged, miscellaneous fairings and antennae were broken, the left wingtip scraped the runway. TESTS AND RESEARCH The nosewheel was sent to the NTSB Materials Lab for examination. No evidence of pre-existing damage or fatigue was found. A video recording of both operating position displays was extracted from the GCS. The videos were spliced together to approximate view from the pilot and sensor operator positions and to compare the relative sight picture between the two cameras. There were some noticeable brief points of data dropouts in the video playback, but analysis by GA-ASI engineers conclude the dropouts were an artifact of the playback unit. The USAF Mishap Analysis and Animation Facility created an animation utilizing the data logger software and on-scene documentation to aid in visualization purposes. The animation includes a chase-plane view of the vehicle and a window of select GCS instrumentation and control positions. The animation was intended for visualization reference use and did not attempt to duplicate the post-impact damage to the vehicle. ORGANIZATIONAL AND MANAGEMENT INFORMATION U.S. Customs and Border Protection is an agency of the Department of Homeland Security. The stated mission of the CBP is to maintain the security of the U.S. border. In 2005, CBP formed the Office of Air and Marine responsible for the aviation assets of the agency, and began developing a UAS program. At the time of the accident, CBP was operating four MQ-9 aircraft from Sierra Vista airport, primarily for training and observation of the U.S.-Mexico border. General Atomics-Aeronautical Systems Inc. is a unit of the General Atomics company. GA-ASI manufactures unmanned aircraft including the MQ-9 Predator-B (Reaper in US Air Force terminology), the MQ-1 Predator-A and others. GA-ASI also manufactures ground control stations, sensor hardware and software, and provides field support services to customers. ADDITIONAL INFORMATION Following the accident, the FAA suspended the CBP Certificate of Authorization (COA) to conduct training on the MQ-9 at Sierra Vista Airport. The CBP conducted an internal review of their programs and oversight, for the purposes of responding to the FAA. On December 17, 2008, the FAA reinstated the training COA and CBP resumed LRE training on February 17, 2009. On October 24, 2007, following a public forum on UAS safety, the NTSB issued recommendations A-07-065 through A-07-069 to the FAA and A-07-070 through A-07-086 to the CBP. At the time of this accident the recommendations were still classified “open-response received.”
the failure of the pilot to timely flare the aircraft to the appropriate attitude, likely associated with the different sight picture after switching cameras, resulting in a bounced landing; and the evaluator pilot’s lack of timely recognition and intervention. Contributing to the accident was the lack of standards and criteria in the Customs and Border Protection initial and recurrent training program for use of the MTS-B camera.
Source: NTSB Aviation Accident Database
Aviation Accidents App
In-Depth Access to Aviation Accident Reports
