CHI96FA211

the pilot's failure to follow the preflight crew briefing and attain a proper altitude during an aerobatic maneuver. Factors in the accident were: the pilot's previous experience of flying similar airshow routines in a different airplane with substantially different performance characteristics, and the company's failure to have a formal training plan for pilots performing airshow demonstration flights.

A McDonnell Douglas Aerospace (MDA) pilot was flying a leased Navy F/A-18C and conducting an airshow practice at St. Louis Regional Airport when the airplane impacted the ground at the bottom of a reverse one-half Cuban eight aerobatic maneuver. The briefed altitude at the top of the maneuver was to be 3,500 feet above ground level (agl), which gave the pilot a base line of 1,000 feet agl. Recorded data showed that the actual altitude at the top of the maneuver was 2,280 feet agl. Using a group of F/A-18 pilots in a F/A-18 simulator, the lowest altitude at the top of the reverse one-half Cuban eight required to successfully complete the maneuver was 2,500 feet agl. The pilot had been trained as a Naval Aviator, and was a graduate of the Navy's test pilot school. He joined MDA Flight Operations 3/4/96. The pilot had accrued 16 hours in the F/A-18 in the last year, of which 11 hours were in the last 90 days. MDA did not have a formal training plan for their pilots who perform airshow demonstration flights. The pilot had flown 5 civilian airshows within the last year; the most recent was 11 days prior to the accident. All the airshows were flown in a Pitts Special.

HISTORY OF FLIGHT On June 19, 1996, at 1448 central daylight time (cdt), a Department of the Navy F/A-18C, Buno Number 165189, leased and operated by McDonnell Douglas Aerospace (MDA), was destroyed after it impacted the terrain while performing a reverse one-half Cuban eight maneuver during a practice airshow at the St. Louis Regional Airport, Alton, Illinois. The commercial pilot sustained fatal injuries. The local 14 CFR Part 91 flight was operating in visual meteorological conditions. No flight plan was on file. The practice airshow flight departed St. Louis Regional Airport, Alton, Illinois, at 1447 cdt. On the morning of the accident, the airplane preflight was initiated by the launching Quality Assurance (QA) inspector. Another QA inspector began to preflight the cockpit and the ejection seat. This inspector was called away to perform a final go/no-go inspection on another airplane that was ready to depart. This inspector stated that while at the other airplane, he received a call, from another McDonnell Douglas inspector, inquiring if he had finished his portion of the preflight. He replied that he still needed to run the seat up and inspect the cockpit lights. By the time this inspector returned to the airplane the pilot had already pulled the ejection seat and canopy pins. The pilot had given the pins to the ground crew to be stored in the 14L door for use upon landing at the St. Louis Regional Airport, if needed. The inspector stated this was not normal procedure and that the pilot should have waited for his return. This procedure is enclosed with the report under the Maintenance Group Chairman's item 14, Aircraft Flight and Inspection Release Form. The preflight was completed and the airplane was taxied to the runway where the go/no-go inspection was performed. The accident occurred during the second demonstration flight for the pilot on June 19, 1996. The first flight on June 19, 1996, was a functional check flight followed by a high altitude practice airshow sequence while en route from the McDonnell Douglas plant (Lambert International Airport, Bridgton, Missouri) to the St. Louis Regional Airport. The airplane departed Lambert Field at 1318 cdt and proceeded to a test area north of St. Louis where the pilot performed some routine inflight systems checks due to recently completed maintenance on the airplane. After performing a partial practice airshow sequence north of Alton, Illinois, the airplane arrived at the St. Louis Regional Airport. The tower cleared the airplane from 500 feet above ground level (agl) to 8,000 feet agl with a 3 mile radius around show center (approximately the center of the airport). The pilot then entered at 1,000 foot agl baseline into the practice airshow routine. The airshow routine, established by MDA in 1993, included the following sequence of events: Takeoff, gear down roll, slow loop, reverse one-half Cuban eight, high speed roll, inverted pass, roll over break, maximum g turn, immelmann, high AOA turn, high AOA pass, high AOA roll, minimum radius 180 degree turn, square loop, barrel roll, and landing. While nearing the top of a loop, the pilot broke off the maneuver because of a cloud that moved over the airport. After a few minutes, the remainder of the practice airshow routine was completed and the airplane landed at 1350 cdt. The airplane was refueled with 6,600 pounds of Jet A. The pilot had invited his family and some friends to watch his practice airshow demonstration and some were in attendance at St. Louis Regional Airport. The pilot met with family and friends both before and after his debrief/brief with the MDA Chief Test Pilot. The second practice flight departed the St. Louis Regional Airport at 1447 cdt. The routine began with a maximum afterburner takeoff, followed by a dirty roll (landing gear extended). After the aircraft completed the roll, the landing gear was retracted and a slow loop was executed. The slow loop was followed by entry to a reverse one-half Cuban eight. The airplane was observed to be low by the Chief Test Pilot who was acting as a safety observer on the St. Louis Regional Airport Air Traffic Control Tower catwalk. He called abort on a hand held radio to the pilot. The transmission of the abort was not acknowledged by the pilot. He did see that the airplane had a positive AOA before impact. Videotape of the accident indicates the airplane impacted the ground at the bottom of the reverse one-half Cuban eight. The time of the impact was approximately 1448 cdt. OTHER DAMAGE Multiple trees, one telephone pole, and a residential garage were damaged during impact sequence. PERSONNEL INFORMATION The pilot was born May 22, 1952. He was the holder of a commercial certificate with single/multi engine land and instrument ratings. The pilot also held a acrobatic competency certificate level two issued on May 20, 1996, with an altitude limitation to 250 feet agl. The acrobatic certificate included the following airplanes; Beech D17S, Beech BE-33C, and a Pitts Special. He held a second class medical issued on June 13, 1996. His most recent biennial flight review was on March 26, 1996. He had accumulated a total of 6,218 hours of flight time. The pilot was trained as a United States Naval Aviator and qualified in the F/A-18. During his military flying career he accumulated approximately 2,255 hours in the F/A-18 airplane prior to coming to MDA Flight Operations in March of 1996. The pilot was also a graduate of the Naval Test Pilot School, located at Naval Air Test Center, Patuxent River, Maryland. The pilot had the following 30/60/90 day flight hour totals in the F/A-18: 3.0/3.9/10.7 hours. The pilot had the following 30/60/90 day simulator hour totals in the F/A-18: 9/27/37 hours. AIRCRAFT INFORMATION The airplane was a McDonnell Douglas F/A-18C, serial number 165189. The airplane had accumulated 20.8 hours time in service at the time of the accident. The engines had 26 hours total hours in service. The most recent continuous inspection was conducted on June 19, 1996. FLIGHT RECORDERS The Deployable Flight Incident Recorder Set (DFIRS) was salvaged from the wreckage. The data was printed out and down loaded prior to NTSB involvement with the accident investigation. The data was loaded into the recovery analysis and presentation system (RAPS) program for visual display of the data collected. In addition, a visual 8MM tape of the right Digital Display Indicator (DDI) showed the Flight Control System (FCS) status display selected with no warnings displayed prior to impact with the terrain. The left 8MM tape was destroyed by the post-crash fire. The DFIRS data was broken down into the four parts of the reverse one-half Cuban eight; Pull-up 180 degree roll Top Bottom 1. Airspeed (knots) 236 176 180 144 2. Baro Alt (ft agl) 1212 1192 2008 472 3. Radar Alt (ft agl) 576 1408 invalid 0 4. AOA (degrees) 25.2 14 12.6 40.6 5. Pitch (degrees) 16.8 58.8 -7.0 22.4 6. Power Lever Angles Left/Right (degrees) 128.1/127.4 full after burner position though out the complete maneuver. The briefed target parameters were; Pull-up 180 degree roll Top Back side 1. Airspeed (knots) 260-280 280-200 200-230 250-270 2. Baro Alt (ft agl) 700-800 2700min 3500min 2000min 3. Radar Alt (ft agl) 700-800 Invalid Invalid Invalid 4. AOA (degrees) 25max 20 0-10 25max 5. Pitch (degrees) 50-55 50-55 0 -90 6. Power Lever Angles Left/right (degrees) 90-102/90-102 Military power position (maximum power needed without afterburner at low airplane weights) See four graphs enclosed with this report. At impact the DFIRS recorded the following throttle and engine parameters; Left Engine Right Engine 1. Power Lever Angle: 128.1 degrees 127.4 degrees 2. High Press Rotor: 16256 RPM 16256 RPM 3. Low Press Rotor: 13568 RPM 13632 RPM 4. Fuel Flow: 9280 Lbs/Hr 9280 Lbs/Hr 5. EGT: 896 Degrees C 896 degrees C 6. Exhaust Nozzle: 68 percent 68 percent At impact the airplane parameter's were recorded by DFIRS; 1. Airspeed: 144 knots 2. Angle of attack: 40.6 degrees 3. Pitch: 22.4 degrees nose up 4. Roll: 9.8 degrees to the left 5. Vertical Velocity: -3840 feet per minute WRECKAGE AND IMPACT INFORMATION The NTSB on-scene investigation began at 0830 on June 24, 1996. There were questions immediately following the accident regarding the ownership of the aircraft and who had responsibility for the investigation. Before the NTSB took over the investigation, a joint investigation between the Naval Safety Center and MDA was in progress. The wreckage had already been removed from the accident site and placed in a MDA hangar. Several inspections of the airplane's components were being performed by the U.S. Navy, MDA and its vendors. The accident site was surveyed by the U. S. Air Force Air Mobility Command, based at Scott Air Force Base, O'Fallon, Illinois. A copy of the survey is attached to this report. The airplane flight path angle at impact was calculated to be minus 16 degrees. This value was calculated from the survey data of the initial impact point and a tree that was struck by the airplane prior to ground impact. The airplane slid between two houses, impacting a telephone pole, several trees and a detached garage structure before breaking up and coming to rest approximately 360 feet from the initial impact point. Evidence of a post crash fire was evident in the general direction of flight from the garage structure forward. Several afterburner flaps were found at the initial impact point, followed in the direction of flight by two distinct furrows corresponding to the outside diameter of the afterburner casings. Scars to either side of these furrows were made by the horizontal stabilators. The airplane's centerline pylon was found between the initial impact point and the garage. The left side leading edge extension was found embedded in the garage structure. The airplane's canopy was recovered 250 feet from the majority of the main wreckage in an area not burned by the ground fire. The canopy unlatch thruster and rocket motors had fired. Most of the glass was broken out in small pieces and scattered over a large area. MEDICAL AND PATHOLOGICAL INFORMATION A post mortem examination of the pilot was conducted on June 20, 1996, at the Madison County Morgue, Edwardsville, Illinois. No pre-existent anomalies were noted during this examination. The pilot's toxicological analysis was performed by both the Federal Aviation Administration (FAA) Civil Medical Institute in Oklahoma City, Oklahoma and the Madison County Corner. The toxicological examination of specimens from the pilot were negative for the drugs scanned. TESTS AND RESEARCH Several F/A-18's pilots were asked to perform the same flight profile as the data obtained from the DFIRS in two F/A-18 simulators. With all aircraft systems operating normally, impact with the ground occurred whenever the top altitude of the Split-S maneuver was less than 2,500 feet agl. In support of this, Split-S maneuvers were flown in the simulator at speeds of 125 to 325 knots, in 25 knot increments, with the top altitude from 3,000 feet agl with 5,500 pounds fuel and max afterburner thrust. Altitude needed to successfully complete this maneuver were constant at 2,500 feet plus or minus 100 feet. When starting the maneuver at 3,500 feet agl, the maneuver could be completed by 1,000 feet agl with altitude available for a smooth transition down to the 500 feet agl minimum. In addition, thrust deficiencies were simulated but, did not replicate data collected from the accident airplane. There were many eyewitnesses to the accident flight. None reported seeing any airplane anomalies. Witnesses were in agreement that the accident airplane took off, performed a slow loop and then initiated a reverse one-half Cuban eight prior to impacting the ground at the bottom of the maneuver. For further information see Operation Group Chairman's report enclosed with this report. According to the Chief Test Pilot, the debrief of the first routine outlined a few constructive comments and finesse techniques on the maneuvers being performed. The comments primarily addressed horizontal maneuvers as all of the vertical maneuvers had been observed to be satisfactory. On one maneuver, the Chief Test Pilot told the accident pilot that he did not like the high speed turn after the inverted pass. The accident pilot agreed stating that he wanted to pull 6.5gs but he only got 5.5 or 6.0gs. The Chief Test Pilot stated that he explained to the pilot that as a result of the turn and the ground track, the next maneuver was rushed. The accident pilot told the Chief Test Pilot that he concurred and explained that it was because he missed the g but it would not be a problem the next time. According to the Chief Test Pilot they then reviewed the parameters for each maneuver. For the reverse one-half Cuban eight, the Chief Test Pilot asked the accident pilot what parameters he was looking for. The accident pilot stated he would be in full burner, looking for 300 (knots) to 320 (knots) going up in the Cuban eight and that he would be playing the back side of the loop with altitude and acceleration. The Chief Test Pilot asked the accident pilot what altitude he would be looking for during the reverse one-half Cuban eight. The accident pilot replied that he would be looking for 3,000 and would pull at 3,500 minimum. The accident pilot told the Chief Test Pilot he would play with the power a little bit depending on his speed and climb angle. He continued to state that he would be pulling 4 to 5 gs on the back side. He then stated that he probably would not get that and he would be switching to 20 to 25 alpha (angle of attack). The Chief Test Pilot questioned the 25 alpha to which the accident pilot responded 20 alpha. The Chief Test Pilot then asked what he would be looking for on the back side. The accident pilot replied that he would probably have to play this to get to his altitude and exit speeds for the rolls and that he would be coming back in with power and accelerating. He stated that he would be playing his altitude to come out at 700 feet (agl) then he would enter the roll. The Chief Test Pilot asked what speed he would be looking for, to which the accident pilot replied he should be around 350 (knots) accelerating close to 400 to do the roll inverted and to brake to the hard turn. According to the Senior Test Pilot who performed previous airshow demonstrations and trained the pilot, he conducted a few one- on-one training sessions with the pilot during which they went over the parameters for each of the maneuvers. The Senior Test Pilot described the parameters shown to the pilot in the company's simulator for the reverse one-half Cuban eight maneuver. The Senior Test Pilot said, as you [the pilot] are completing the back side of the slow loop maneuver, "...Normally you have to play altitude down; there was never a problem with being low on that maneuver. You would normally pull out at 1000 to 900 feet agl and have the airplane fly down to five hundred feet while accelerating to set up for the reverse half Cuban eight. The acceleration takes airspeed up to 260, 270, 280 knot area. That's with mil power, you don't need more than that. If you needed a little bit more, after burner would shoot you up to 300 knots. At that point, use a pretty smart pull to set the attitude for reverse one-half Cuban eight. Looking for 50 to 55 degrees but could go as high as 65. You may be down as low as 45. The range here would be where you are accommodating different winds; headwinds or tailwinds. You are looking to get distance so that you can come down the back side of that and be able to get speed up for the rolling sequence which follows. Pulling the aircraft up, set the attitude again, or the flight path at this point since angle of attack is relatively low. Just watching altitude now. Altitude, I use 2700 to 2800 feet agl as the point to roll the airplane inverted and then just pause there because the maneuver looks better rather than just a roll and pull and then just extend on up. Plus you are looking for 3500 feet agl. That was the constant minimum that we used for these two overhead maneuvers." The Senior Test Pilot said that he never really considered the reverse one-half Cuban eight maneuver to be a particularly challenging one nor the rest of the maneuvers used in the airshow routine. Initial on-scene inspection of the engines indicated the left engine showed less rotational damage than the right engine, which showed high rotational damage. The left engine was selected for shipment to Naval Aviation Depot (NADEP) Jacksonville for teardown and detailed investigation under the supervision of the NTSB. In addition, the left variable exhaust nozzle (VEN) assembly, both high pressure compressors (HPC) variable geometry actuators, the fan variable geometry actuator, the power lever control and throttle box, the VEN position transmitter and the main and afterburner fuel pumps and controls, were also sent to NADEP Jacksonville for teardown and investigation. The right engine VEN assembly, HPC and fan variable geometry actuators, power lever control and throttle box and main fuel control, the VEN position transmitter and the main fuel control and pump were also selected for shipment to NADEP Jacksonville. Based on all evidence examined on-scene and at NADEP Jacksonville, the right engine was operating at or near the maximum afterburner power setting at ground impact. The rotating core exhibited evidence of high rotational energy at ground impact. The engine control components examined all exhibited impact marks indicative of a high (nearly maximum afterburner) power setting. The VEN components all consistently indicated a maximum afterburner nozzle area position. Charred wood pressed into the aft end of the right engine and the soft deformation of the exhaust centerbody indicated combustion heat was present in the turbine section during the impact sequence. The physical evidence exhibited by the right engine and its components supported the pre-impact operational data downloaded from the airplane's data recording system. Initial inspection revealed little evidence of high rotational energy in the left engine; however, a close inspection of the left engine did reveal signs of high rotational energy at some point during the accident sequence. There was evidence of rubbing between rotors and stators, stator cases and static seals to indicate that the engine was rotating at initial impact. The left engine control components examined all exhibited impact marks indicative of a high (nearly maximum afterburner) power setting and VEN components all consistently indicated a maximum afterburner nozzle area position at impact. Charred wood found in the combustion case indicated that heat was present in the combustor when the debris was ingested. Based on the evidence found during the detailed disassembly, and the correlation between the left and the right VEN positions at impact, it was concluded that the left engine was operating at or near maximum afterburner power at impact. Examination of the right and left engine indicated normal operation throughout the flight. For further information see Propulsion Group Chairman report enclosed with this report. Maintenance records revealed significant maintenance conducted on the airplane prior to the accident. Fuel tank number 1 was disassembled to inspect for suspect cracked clips on the left and right fuselage structure. The clips were replaced and tank 1 was reassembled. Fuel tank number 4 was disassembled to look for the cause of a re-occurring cavity drain leak. No obvious leak source was discovered, so the bladder tank was removed and replaced. The left and right generators were removed and reinstalled to facilitate the troubleshooting of the fuel tank number 4. The fuel tank number 1 fuel quantity probe was removed and replaced for a fuel quantity indication problem. The right motive flow fuel "T" in door 53R was disassembled for motive flow fuel leak. An O-ring was replaced and the fuel coupling was reassembled. A complete functional check of the fuel system was satisfactorily completed during a post maintenance ground engine run. This work was done by non-union MDA personnel due to a strike of the International Association of Machinists on the property of MDA. Examination of the fuel system indicated normal fuel system operation and no anomalies were discovered. For further information see Maintenance Records Group Chairman's report enclosed with the report. The aircrew ejection seat components recovered indicated partial/incomplete firing of pyrotechnic items. Most items appeared to have been expended as a result of exposure to extreme heat and/or post crash fire. The ejection seat catapult did not activate. The ejection seat did not leave the cockpit. The Seat Firing Handle was withdrawn from the position it would be installed in during normal flight operations. The handle was in a position forward of and below its normal position when installed in the seat bucket forward beam. The interface link between the handle assembly and the seat initiator firing handle sear assemble was bent approximately 110 degrees from the vertical position over the forward edge of the handle receiver block. A witness mark was noted on the upper forward surface of the handle receptacle in the receiver block which matches the physical characteristics of the handle interface link. The Mechanically Activated Initiator, which provides manual canopy jettison capability, was recovered from the airplane wreckage. The output Shielded Mild Detonating Cord (SMDC) line has been expended and appeared to have functional normally. Visual examination of the inner diameter of the SMDC line, which was sheared off at the surface of the line installation nut, contained evidence the explosive core sear rod was withdrawn fully from the upper end of the unit and was not recovered. The metal guard which normally surrounds the area occupied by the sear rod was mechanically damaged. The actuation sear rod is absent from the unit with no witness marks or damage in the opening where the sear rod is normally installed. For further information see Engineering Investigation of Ejection and Aircrew Escape System Report enclosed with this report. The IIC calculated, based on ejection seat performance data from Naval Air Training and Operating Procedures (NATOPS), and the accident airplane performance parameters as recorded on DFIRS, the pilot's lowest altitude to safely eject would have been approximately 320 feet agl or two seconds before impact. The video tape recording system (VTRS) tape of the right DDI Flight Control system (FCS) page was recovered. The FCS page recording contained useful data until approximately 4 seconds before electrical power was lost at impact. The Head Up Display (HUD) VTRS was recovered but the tape was destroyed by the post-crash fire. No faults or flight control cautions were observed in reviewing the VTRS recording of the flight control page (FCP) in the right DDI. Laboratory X-ray, wire ring-out, and functional tests of the flight control actuators (stabilator, aileron and rudder) indicated no anomalies. Examination of the leading edge flap (LEF) overtravel stops indicated the right wing LEF was 13 degrees down and the left wing LEF was 34 degrees down. Deployable Flight Incident Recorder Set (DFIRS) data indicated both left and right wing (inboard and outboard) LEF were fully extended at 32.5 degrees down. The video recording of the FCS page also indicated near max LEF extension (33 degrees) on both left and right LEF's. The DFIRS did not deploy from the airplane and subsequently received extensive burn damage. The damage was too extensive to allow normal readout of the data at the MDA facilities. The DFIRS data was recovered by sending the damaged unit to the manufacturer for removal of memory chips and then returning the memory chips to MDA for readout. The airplane's data storage unit (DSU) was recovered and returned to the MDA facility for data recovery. Data was successfully read out from the airplane DFIRS and DSU. DFIRS data were recorded to within .5 to 1.5 seconds of electrical power loss at impact. The electronic data from the DFIRS, DSU and DDI FCP showed normal flight control operation with no faults or failures detected. For further information see Systems Group Chairman's report enclosed with this report. The Defense Logistics Agency Contractor's Flight and Ground Operations manual specifies MDA Pilot minimum currency requirements of 35 hours flight time in the previous six months, and allows that fifty percent of the flight time required can be simulation time. The pilot last flew an F/A-18, 19 days prior to the accident date. The pilot also flew five civilian airshows in a Pitts Special, within the last year, with the last airshow on June 8, 1996, 11 days before the accident date. For further information see Operations report enclosed with this report. ADDITIONAL DATA Parties to the investigation were the Federal Aviation Administration; MDA; Naval Safety Center; International Association of Machinists; and General Electric. Following the on-scene portion of the investigation, the wreckage was released to a MDA representative on July 30, 1996. *Due to limitations within the computer system the last digit of the aircraft Buno number could not be added to the Registration Number under Aircraft Information. -