Cirrus Aircraft SR22

Full Narrative

HISTORY OF FLIGHTOn September 1, 2022, about 1707 central daylight time, a Cirrus Aircraft SR22 airplane, N420SS, was substantially damaged when it was involved in an accident near Tomball, Texas. The flight instructor was fatally injured; the pilot and passenger sustained serious injuries. The airplane was operated as a Title 14 Code of Federal Regulations Part 91 instructional flight.

Earlier in the week, on August 29, 2022, the pilot accepted delivery of his factory-new Cirrus SR22 airplane at the Cirrus Aircraft Vision Center located at the McGhee Tyson Airport (TYS), near Knoxville, Tennessee. He had no previous flying experience in a Cirrus airplane besides a 30-minute demonstration flight that was flown about 1.5 years before the accident. As part of his purchase agreement, the pilot was scheduled to receive transition training in his new airplane for the remainder of the week. The pilot’s first interaction with his assigned Cirrus Aircraft Factory Flight Instructor was the night before he accepted delivery of his airplane. During the phone call, the flight instructor asked what the pilot wanted to accomplish during his flight training, and the pilot replied that he needed to learn how to fly instrument approaches and to make takeoff and landings in the airplane.

According to the pilot, his first flight in the airplane consisted of a takeoff, traffic pattern, and landing. No flight instruction was provided during this initial 10 to15 minute delivery flight that was completed with another Cirrus Aircraft pilot. After the pilot accepted delivery of his airplane, the assigned flight instructor began the pilot’s transition training. The first task completed was to review operating parameters for and demonstrate how to use the Cirrus Airframe Parachute System (CAPS) using a simulator. Afterwards, they reviewed the airplane’s fuel and electrical systems on a whiteboard and introduced the Cirrus Perspective+ integrated avionics system using a tabletop simulator.

On August 30, 2022, the pilot and his flight instructor began flight training in his airplane. Before their flight they discussed how to preflight the airplane. Due to engine power restrictions (maintain at least 75% power) for a new airplane, they were unable to conduct traffic pattern work and the pilot felt they “couldn’t really learn anything” that he hoped to achieve on the flight. Instead, they focused on using the Cirrus Perspective+ system. The pilot stated that the flight instructor did not provide more feedback on how to fly the airplane, such as providing the different airspeeds to be flown during the different phases of flight. When interviewed, the pilot expressed his concern that Cirrus Aircraft “didn’t have some syllabus for me” dictating what he was to learn each day.

On August 31, 2022, the weather at TYS was windy, so the pilot asked his flight instructor if they could fly to an uncontrolled airport where they could work on flying the airplane in the traffic pattern and practice takeoff and landings. According to the pilot, it was during this flight that the flight instructor first provided the reference airspeeds for downwind, base, and final approach. The pilot stated that he placed a note with the reference airspeeds on the cockpit dashboard.

The pilot stated that before the fourth day of transition training, he woke up overnight shivering and sweating. On the morning of September 1, 2022, the pilot told his flight instructor that he did not feel well, and the decision was made to fly to the pilot’s homebase located at David Wayne Hooks Memorial Airport (DWH), Spring, Texas. Based on interviews, it was not clear with whom the pilot would have continued his transition training after returning to his homebase. Cirrus Aircraft’s Chief Pilot stated the pilot would have likely resumed training with a local Cirrus Standardized Instructor Pilot (CSIP), while the pilot was under the impression that he would have continued with the factory flight instructor if he was feeling better the next day. The flight instructor had not planned on flying to DWH and, as such, he had to reschedule some other work obligations and kennel his dog for at least one night. The pilot stated that before they departed on the cross-country flight, he and the instructor did not communicate what roles each would have should an emergency arise during the flight.

The first flight leg was supposed to be from TYS to Alexandria International Airport (AEX), Alexandria, Louisiana. However, due to adverse weather that impacted their intended route, the flight diverted to Monroe Regional Airport (MLU), Monroe, Louisiana. The pilot stated that the flight instructor appeared to briefly fall asleep during the flight from TYS to MLU. The pilot stated that he flew most of the flight with the airplane’s automatic flight control system (AFCS) engaged, making necessary heading changes using the heading bug on the primary flight display (PFD).

The pilot stated the airplane was topped-off with fuel after landing at MLU and that they were on the ground for about 30-45 minutes, during which the pilot and the passenger each drank a cup of coffee, and the flight instructor drank a soda. The pilot stated that shortly after they departed MLU, about 10 minutes into the flight, the flight instructor told him that he needed to urinate. The pilot offered the flight instructor one of his “Little John” pilot urinals, but the flight instructor declined to use the urinal. The pilot stated that the flight instructor appeared to be in discomfort (shifting around in his seat and grimacing) for the remainder of the flight and did not speak much or provide any feedback until they got closer to DWH.

The pilot stated that the airplane’s AFCS was engaged for most of the flight from MLU to DWH, and that he used the heading bug on the PFD to make any heading changes that were issued by air traffic control. The flight was on an instrument flight rules flight plan, and the controller issued several vectors to keep the airplane clear from areas of adverse weather. As the flight approached DWH, the pilot listened to the Automatic Terminal Information Service broadcast and selected the RNAV runway 17R instrument approach at DWH using the Cirrus Perspective+ system, but he was unsure if he activated the approach.

While the flight tracked north toward Conroe, Texas, the controller asked if they wanted the full RNAV runway 17R approach or the visual approach to runway 17R. The flight instructor replied to the controller that he wanted the visual approach to runway 17R. The pilot told his flight instructor that he had never flown a visual approach before and asked how to use the Cirrus Perspective+ system during this type of approach. The flight instructor then showed the pilot how to “scroll-down” on the display to see data associated with a visual approach. When interviewed, the pilot stated that he did not know how the visual approach was supposed to work in the Cirrus Perspective+ system and that he was confused that there was no altitude step downs or waypoints visible after the visual approach was selected. The controller issued a heading to intercept the final approach course to runway 17R at DWH, cleared the flight for the visual approach, and told the pilots to contact the DWH tower controller.

The pilot stated that he saw the runway and its associated precision approach path indicators (PAPI) lights after the airplane turned onto the final approach course and that the airplane appeared to be on a proper descent path to the runway. The airplane’s airspeed began to decrease as the flight continued toward the runway, and the flight instructor told him to “give it some throttle” to increase airspeed. The pilot increased the throttle slightly but noted that he did not hear the engine “roar” with power. The flight instructor stated “My airplane” or “I’ve got the controls” shortly after the pilot increased the throttle. The pilot estimated “a few seconds” transpired between his increase of throttle and when the flight instructor took control of the airplane.

The pilot stated that after the flight instructor took control of the airplane, the airplane descended below the proper glidepath where he could no longer see the PAPI system or the runway. The pilot stated that in the moments before the accident the flight instructor rolled the airplane into a left-wing-down attitude, likely trying to maneuver the airplane into a clearing left of the airplane’s position. The airplane impacted several trees before it came to rest in a wooded mobile home neighborhood.

When asked, the pilot did not recall completing the pre-landing checklist but stated that he believed it was something that would have been completed before the accident. Additionally, he could not specifically recall individual positions of the throttle, mixture control, and fuel selector at the time of the accident; however, he recalled the ignition/magneto switch was positioned to both. The pilot stated that he believed the airplane’s automatic flight control system (AFCS) was still engaged when the flight instructor took control of the airplane. The pilot stated that the flight instructor did not ask him to verify control positions or troubleshoot anything in the moments before the accident, nor did they discuss any anomalies with the airplane or if they should deploy the CAPS. The pilot stated that he believed the engine was operating at the time of the accident, but thought it was odd that he did not hear the engine “roar” with power after the flight instructor took control of the airplane and increased the throttle. The pilot, flight instructor, and passenger were wearing noise-canceling headsets during the flight.

The pilot did not recall completing any emergency action procedures before impact. Further, the pilot reported that he did not interfere with, nor did he remember the instructor interfering with the fuel mixture control throughout the latter duration of the flight. He stated that a tablet was strapped to his knee and a cell phone was in the seat pocket. Likewise, according to the pilot, the flight instructor did not have any items that could interfere with the operation of the airplane.

A review of recorded data downloaded from both the airplane’s Cirrus Perspective+ system and Recoverable Data Module, along with recorded ATC communications, revealed that the flight received radar vectors to join the visual approach to runway 17R at DWH, as shown in Figure 1. At 1701:36, the flight was cleared for the visual approach and told to contact the DWH tower controller. At 1703:51, the tower controller cleared the flight to land on runway 17R. The airplane’s AFCS was engaged during the earlier portions of the approach, and the flight crew used vertical speed mode to descend to and level off at specified intermediate altitudes based on a lateral navigation only (LNAV) instrument procedure, as shown in Figure 2 and Figure 3. Additionally, according to the recorded data, there were several engine power changes (increases/decreases) made during the approach.


Figure 1 – Plot of the airplane ground track.


Figure 2 – Plot of airplane altitude, ground speed, true airspeed, indicated airspeed, and vertical speed.

Figure 3 – Plot the airplane’s descent profile during the visual approach.

Based on recorded data from the airplane, beginning at 1705:58, there was a loss of fuel flow to the engine, as shown in Figure 4, as the airplane continued in a descent toward runway 17R in visual meteorological conditions, as shown in Figure 5. At that time, the airplane was about 1,126 ft msl (988 ft agl) and 2.26 nm from the runway 17R displaced threshold. About 16 seconds after the loss of engine power, the AFCS was turned off and remained off for the remainder of the flight. With the AFCS disengaged, the flight crew continued the descent toward the runway under manual flight control.


Figure 4 – Plot of engine speed, EGT, engine power, fuel flow, and manifold pressure.

Figure 5 – Plot of the airplane’s descent path at the end of the flight.

Based on recorded engine manifold pressure values after the total loss of fuel flow to the engine, the power (throttle) lever was decreased once and then increased twice. The first, partial, increase of throttle began at 1706:29 as the airplane descended through 738 ft msl (602 ft agl) about 1.49 nm from the runway 17R displaced threshold. About 11 seconds later, there was a second throttle increase, likely to the full throttle position, as the airplane descended through 544 ft msl (407 ft agl) about 1.23 nm from the displaced threshold.

According to the recorded data, at 1706:50, about 394 ft msl (258 ft agl) and 1.01 nm from the displaced threshold, the wing flaps were retracted from 50% to 0%, as shown in Figure 2. About 6 seconds later, the stall warning tone activated as the airplane decelerated below 82 knots indicated airspeed (KIAS). At 1707:00, the final recorded data, indicated that the airplane had descended to 214 ft msl (76 ft agl) and decelerated to 74 KIAS. According to the Cirrus SR22 Pilot Operating Handbook (POH), the airplane’s wings-level aerodynamic stall at maximum gross weight is 74 KIAS. The airplane subsequently impacted trees and terrain about 0.76 nm from the runway 17R displaced threshold. PERSONNEL INFORMATIONFlight Instructor

A comprehensive flight record for the flight instructor was not located during the investigation. On March 17, 2022, the flight instructor reported 697 total flight hours when he submitted his application for employment with Cirrus Aircraft. Based on Cirrus Aircraft’s flight records, on April 20, 2022, the flight instructor flew 1.2 hours in a Cirrus SR22 airplane in conjunction with his interview for a Cirrus Factory Flight Instructor position. Besides the interview flight, the flight instructor did not have any flight time in a Cirrus airplane nor was he qualified as a Cirrus Standardized Instructor Pilot (CSIP) before he was hired by Cirrus Aircraft in May 2022.

While employed by Cirrus Aircraft, the flight instructor flew 109.4 hours, of which 44.9 hours were dual-instruction-received in conjunction with his Cirrus SR22 transition training and CSIP qualification. On July 1, 2022, the flight instructor received his CSIP qualification following a successful 1.7-hour checkride administered by another Cirrus Factory Flight Instructor. While employed by Cirrus Aircraft, the flight instructor flew 14.2 hours as pilot-in-command in support of internal company flight operations and provided 48.6 hours of dual-flight-instruction to four customers, including the accident pilot.

The investigative team interviewed several factory flight instructors who were responsible for training the accident flight instructor to perform his duties with Cirrus Aircraft’s customers. The training included three phases; transition training in the Cirrus SR22, an intermediate phase to gain additional flight experience in the SR20/22 airplane while supporting internal company flight operations, and finally the completion of CSIP training.

The primary instructor pilot responsible for the training noted that the accident flight instructor had difficulties in progressing through the Cirrus SR22 transition training program. Documented in training records were concerns about the accident flight instructor maintaining situation awareness while operating in areas of high traffic (Class B airspace), maintaining positive control (i.e. “staying ahead”) of the airplane in dynamic situations, and decision-making regarding CAPS deployment. The primary instructor pilot stated that after 23.9 hours of dual instruction completed over a 9-day period, he felt the accident flight instructor was not ready to fly solo and required additional training once a relocation to the Cirrus Aircraft Vision Center in Knoxville, Tennessee, could occur and before the trainee would be ready to transition to the company’s CSIP training. The accident flight instructor’s training records also listed difficulties with managing airplane automation and airplane handing. The consensus from the instructors interviewed was that while the accident flight instructor did require additional time and performed slightly below average in comparison to other new hires in training, he did ultimately perform to a satisfactory level and was proficient enough to work with customers as a Cirrus Factory Flight Instructor.

Pilot

The pilot was unable to provide a comprehensive record of his flight experience. The pilot’s current flight logbook reflected his total time flown in his Cessna 182 that he purchased new in September 2005. As of August 20, 2022, he had flown 1,686.2 hours in his Cessna 182. The pilot reported additional flight time in a Pilatus PC-12 and a Cessna 414, but most of his pilot-in-command flight experience was flown in his Cessna 182. On his last Federal Aviation Administration (FAA) medical application, dated March 8, 2022, the pilot reported 1,535 hours total flight time, of which 82 hours were flown in the last 6 months.

According to a pilot history questionnaire he provided to Cirrus Aircraft before beginning his Cirrus SR22 Transition Training, the pilot had flown about 2,039 hours, of which 1,925 hours were flown as pilot-in-command. Besides a 30-minute demonstration flight completed about 1.5 years before the accident, the pilot did not have any flight time in a Cirrus airplane before he accepted delivery of his new airplane. According to his flight logbook and recorded data from the airplane, the pilot had received 12 hours of dual-flight-instruction in the accident airplane at the time of the accident.

Before he arrived at the Cirrus Aircraft Vision Center Campus in Knoxville, Tennessee, the pilot communicated with the person responsible for his airplane delivery experience, and the person responsible for scheduling his training. The pilot stated that he was unable to access the Cirrus Approach Learning Portal before he traveled to Knoxville to accept delivery of his airplane. He reported his inability to access the training portal to the training manager who told the investigation team in a subsequent interview that the issue was elevated to Cirrus Aircraft’s IT department for resolution; however, for unknown reasons, the pilot remained unable to access the training portal. As an interim solution, before he traveled to accept delivery of his airplane, the pilot contacted his Cirrus Aircraft sales representative who provided a copy of the airplane’s POH. The pilot was already in the Knoxville area and began his transition training when his access issue to Cirrus Approach Learning Portal was finally resolved.

When interviewed, the pilot repeatedly stated that he had no experience with Cirrus airplanes and believed Cirrus Aircraft would provide a syllabus for his transition training. The pilot stated that, given his lack of flight experience in the Cirrus SR22, his training expectations were to be taught how to fly his new airplane and, as such, was “relying on them to teach me, not me [to] tell them what to teach me.” AIRCRAFT INFORMATIONA postaccident review of the available maintenance records found no unresolved airworthiness issues.

According to the Cirrus SR22 POH, the airplane’s maximum glide ratio was 8.8 to 1. During a forced landing with no wind, the airplane flown at 92 KIAS with the flaps fully retracted would glide 1.45 nm laterally for every 1,000 ft of altitude loss. The POH does not provide glide performance data for the airplane with flaps extended to 50% or 100%.

The Cirrus Perspective+ integrated avionics system, built by Garmin for Cirrus Aircraft, presents flight instrumentation, position, navigation, communication, and identification information to the pilot through large-format displays. The automatic flight control system (AFCS) provides the flight director, autopilot, yaw damper, and manual electric trim functions. The system features two 12-inch, high resolution display units. The left display is configured as a primary flight display (PFD). The right display is configured as a multi-function display (MFD). AIRPORT INFORMATIONA postaccident review of the available maintenance records found no unresolved airworthiness issues.

According to the Cirrus SR22 POH, the airplane’s maximum glide ratio was 8.8 to 1. During a forced landing with no wind, the airplane flown at 92 KIAS with the flaps fully retracted would glide 1.45 nm laterally for every 1,000 ft of altitude loss. The POH does not provide glide performance data for the airplane with flaps extended to 50% or 100%.

The Cirrus Perspective+ integrated avionics system, built by Garmin for Cirrus Aircraft, presents flight instrumentation, position, navigation, communication, and identification information to the pilot through large-format displays. The automatic flight control system (AFCS) provides the flight director, autopilot, yaw damper, and manual electric trim functions. The system features two 12-inch, high resolution display units. The left display is configured as a primary flight display (PFD). The right display is configured as a multi-function display (MFD). WRECKAGE AND IMPACT INFORMATIONThe accident site was in a wooded mobile home neighborhood about 0.76 nm north (352° true) of the runway 17R displaced threshold. The initial impact point was a group of 60 to75-foot-tall pine trees. Multiple pine trees were knocked over by the airplane and there were numerous tree branches scattered along the wreckage debris path and among the main wreckage. The CAPS was found deployed at the accident site and is further discussed in the Survival Aspects section.

The main wreckage consisted of the cabin, left wing, aft fuselage, empennage, engine, and propeller. The right wing separated from the fuselage at the wing root and was found on the opposite side yard of the mobile home structure. The left aileron was found adjacent to the right wing. All remaining flight controls (right aileron, right flap, left flap, elevator, and rudder) remained attached to their respective support hinges.

Flight control continuity for the elevator, rudder, and aileron could not be established due to impact damage; however, the observed cable separations were consistent with overstress. The roll and pitch trim motors were found in neutral trim positions. The flap selector was found in the UP position. The wing flap actuator jack screw was found fully extended, consistent with fully retracted flaps.

Before wreckage recovery, the power (throttle) and mixture controls were found in the full forward position. The 3-position switch for the electric fuel pump was in the OFF position. The engine starter/ignition key switch was positioned to the left magneto. The fuel selector lever was found positioned to the left fuel tank. An examination of the fuel selector confirmed the valve was positioned to the left fuel tank at impact. There was a strong odor of 100-low lead aviation fuel at the accident site. Neither fuel tank contained any measurable amount of fuel; however, based on recorded data there was ample fuel in both fuel tanks upon impact.

The entire fuel system was examined from each wing tank to the individual engine fuel injectors. There was no evidence of any preimpact restrictions or debris in the fuel system. The fuel tanks were clean and without any notable debris. The fuel tank venting was unobstructed and revealed no anomalies when air was blown through the vent lines. The fuel check/flapper valve located in each wing’s fuel collector tank functioned normally. All fuel supply and return lines were fractured near their respective wing roots. The fuel supply and return lines from the fuel selector to the firewall fittings were free of obstructions and debris when air was blown through the lines.

The flexible fuel supply line from the firewall to the electric fuel pump inlet was clear of obstructions and debris. The electric fuel pump functioned when connected to the airplane’s 28-volt battery. The 90° elbow inlet fitting to the fuel supply line to the electric fuel pump was found separated. There was impact-related damage to the lower right engine compartment near the electric fuel pump. Blue fuel staining was observed on the interior and exterior surfaces of the inlet fuel fitting. The 90° elbow inlet fitting and fuel supply line were submitted to the NTSB Materials Laboratory for additional examination to determine whether the fitting separated while inflight or during impact and is discussed further in Research and Testing.

The fuel line from the electric fuel pump to the fuel filter assembly was intact and the B-nut fittings were tight. There were no obstructions, restrictions, or debris found in the fuel line between the electric fuel pump and the fuel filter assembly. The fuel filter assembly screen was clean and free of debris.

The fuel line from the fuel filter assembly to the engine-driven fuel pump was intact and the B-nut fittings were tight. There were no obstructions, restrictions, or debris found in the fuel line between the fuel filter assembly and the engine-driven fuel pump. There was a small amount of clean 100 low-lead aviation fuel drained from the fuel line to the inlet side of the engine-driven fuel pump, and from within the engine-driven fuel pump. The engine-driven fuel pump drive coupling was intact, and the engine-driven fuel pump functioned when tested. The fuel line between the engine-driven fuel pump and the fuel flow transducer contained a small amount of clean 100-low lead aviation fuel. The fuel transducer was clear of any debris and air passed freely through the assembly. The fuel lines downstream of the fuel transducer contained clean 100 low-lead aviation fuel.

The engine remained attached to its engine mount and the firewall. There was mechanical continuity between the throttle and mixture controls to their respective engine components. The upper spark plugs were removed and exhibited features consistent with normal engine operation. Internal engine and valve train continuity were confirmed as the crankshaft was rotated through the crankshaft flange. Compression and suction were noted on all six cylinders in conjunction with crankshaft rotation. A borescope inspection of each cylinder did not reveal any anomalies with the cylinders, pistons, valves, valve seats, or lower spark plugs. Both magnetos remained attached to their engine installation points and provided spark on all ignition leads in conjunction with crankshaft rotation. Compressed air was applied to the fuel line downstream of the fuel flow transducer to test the fuel manifold. The compressed air discharged fuel and air from all 6 fuel injector lines. The vapor return line from the fuel manifold to the engine-driven fuel pump was free of obstructions when air was blown through the line.

Additional examination and bench testing of the engine-driven fuel pump, fuel flow transducer, throttle and control unit, fuel distribution valve, fuel injector lines, and fuel injector nozzles did not reveal any preaccident mechanical discrepancies that would have precluded normal operation. ADDITIONAL INFORMATIONCirrus Airframe Parachute System (CAPS) Deployment Guidance

In the event of a loss of engine power, the use of CAPS for Cirrus SR22/SR22T (G5 and subsequent models) is as follows:

- Below 600 ft agl, do not deploy CAPS and land the airplane straight ahead. According to the Cirrus Aircraft SR22 POH, altitude loss from level flight deployments has been demonstrated at less than 400 ft agl; however, deployment at such a low altitude leaves little or no time for the airplane to stabilize under the canopy or for the cabin to be secured. A low altitude deployment increases the risk of injury or death and should be avoided. The chance of a successful deployment greatly decreases below 400 ft agl.

- If the loss of engine power occurs between 600 ft agl and 2,000 ft agl, deploy CAPS immediately.

- If the loss of engine power occurs above 2,000 ft agl, the pilot can troubleshoot the loss of engine power and then deploy CAPS if still required.

- During an engine failure that occurs above 2,000 ft agl, where there is no airport within gliding distance, the pilot should troubleshoot the loss of engine power and maneuver to the best location for CAPS deployment. Upon reaching 2,000 ft agl, if not already deployed, CAPS deployment is recommended.

- If an engine failure occurs within gliding distance of a runway, the pilot must continually monitor the situation to verify if the landing is assured. At 2,000 ft agl, the pilot must reassess if the landing is assured and, if not, then deploy CAPS. At 1,000 ft agl, reassess if the landing is assured and, if not, then deploy CAPS.

An immediate CAPS deployment is required if the airplane enters an aerodynamic spin at any altitude. The immediate deployment of CAPS is recommended if there is a loss of airplane control, structural failure, or if no other survivable alternative exists.

During interviews of Cirrus Aircraft employees who had provided and/or supervised transition training, there was some variability in how they perceived under what parameters the CAPS should be used. While none of the interviewees gave an answer below the minimum established parameters for CAPS deployment, several factory flight instructors varied on how long they believed they would troubleshoot a loss of engine power, at what altitude troubleshooting was no longer an option, and whether they were willing to attempt a glide to a surface suitable for landing (including the runway while in the traffic pattern).

The CAPS training model reinforced that anyone can deploy CAPS should they believe the situation calls for it. This rationale allows for an adequately briefed passenger to identify when and how to deploy CAPS in case of a pilot incapacitation emergency.

Pilot-In-Command (PIC) During Flight

Under federal regulations, the pilot was appropriately certificated to act as PIC during his transition training and the accident flight. The pilot reported feeling unwell on the morning of the flight. FAA Aeronautical Information Manual Chapter 8, Section 1, Fitness for Flight, states that:

Even a minor illness suffered in day-to-day living can seriously degrade performance of many piloting tasks vital to safe fight. Illness can produce fever and distracting symptoms that can impair judgment, memory, alertness, and the ability to make calculations. Although symptoms from an illness may be under adequate control with a medication, the medication itself may decrease pilot performance.

The safest rule is not to fly while suffering from any illness. If this rule is considered too stringent for a particular illness, the pilot should contact an aviation medical examiner for advice.

When interviewed, the pilot believed his instructor was the PIC for the flight although he acknowledged they never had a conversation to confirm that presumption. Additionally, the pilot stated that before they departed on the flight, he and the instructor did not communicate what roles each would have should an emergency arise.

For purposes of insurance coverage for the accident airplane, to act as PIC, the policy required the pilot to complete an instrument proficiency checkride or “flight school” in the make/model within 12 months before the intended flight. The pilot’s insurance policy did not specify a specific flight hour requirement in make/model before the pilot could act as PIC. Based on the available documentation, the pilot had not completed his factory transition training, nor had he completed an instrument proficiency flight in a Cirrus SR22 airplane.

During interviews of the Cirrus Aircraft employees who provided and/or supervised transition training, there was a difference in opinion about who acts as PIC during customer transition training and/or repositioning flights. When asked, most of the factory flight instructors or supervisors stated that if the customer had not completed standardization training, the factory flight instructor would act as PIC and would have ultimate responsibility for the safety of the flight (including emergency situations requiring CAPS deployment). However, notably, the Cirrus Aircraft’s Chief Pilot disagreed with this assertion, stating that if the customer was an appropriately certificated pilot they could act as PIC without having completed their standardization training. MEDICAL AND PATHOLOGICAL INFORMATIONFlight Instructor

The flight instructor experienced cardiac arrest while being treated by emergency medical services (EMS). He was transported to a hospital, but resuscitation efforts were unsuccessful and he was declared dead by emergency department (ED) personnel.

According to the flight instructor’s autopsy, conducted by the Montgomery County Forensic Services Department, the cause of death was multiple blunt force injuries, and the manner of death was accident. The flight instructor sustained injuries to the head, chest, ribs, pelvis, legs, and lumbar vertebra. The urinary bladder was lacerated and empty. The autopsy did not identify significant natural disease.

At the request of the Montgomery County Forensic Services Department, NMS Labs performed postmortem toxicological testing of femoral blood from the flight instructor. No tested-for substances were detected. The FAA Forensic Sciences Laboratory also performed postmortem toxicological testing of specimens from the flight instructor and no tested-for substances were detected.

Pilot

The pilot was alert and walking when EMS personnel arrived. He was transported via ambulance to a hospital ED for evaluation of his injuries where imaging studies showed no significant natural disease. He tested positive for COVID-19 while in the hospital ED. However, additional hospital laboratory studies were otherwise generally unremarkable but did not include testing for alcohol or other drugs. A review of additional medical documentation revealed that the pilot sustained sensorimotor polyneuropathy (serious injury), as well as a laceration to the posterior head and general abrasions to the body and face. SURVIVAL ASPECTSCabin Seating

The airplane was equipped with five seat positions, two forward-facing crew seats and a single bench seat with three forward-facing passenger seats positions in the aft portion of the airplane cabin. During the flight, the flight instructor was seated in the right crew seat, the pilot in the left crew seat, and the passenger in the rear center seat position.

The right crew seat was found attached to the seat rail but separated from the fuselage and upright a short distance away from the main wreckage. The upper and lower seat pans, energy absorption module, and front cross tubes were damaged, with crushing, dents, and tears in multiple places.

The left crew seat remained attached to fuselage structure; however, the seat separated during the wreckage recovery. The rear seat, consisting of the bench and seat backs, remained attached to fuselage structure in the aft portion of the cabin.

Restraints

The crew seats were equipped with an AmSafe lift latch buckle four-point restraint system with inertia reel and airbag in the outboard shoulder harness. The airbags deployed in the accident. The rear seats were equipped with a push-button buckle three-point restraint system and inertia reel, but no airbags.

The pilot stated that his restraint was fastened during the accident, and he was able to release the latch to evacuate after the accident. The surviving passenger stated her restraint was also fastened and held her in place until the pilot came, supported and unbuckled her. According to first responders, the flight instructor separated from his crew seat and was found on a nearby covered porch with railing.

When investigators arrived at the accident site both crew seats, all restraints were unbuckled except the right rear seat. All five inertia reels functioned as expected (locked) when rapidly pulled during postaccident testing.

On April 24, 2023, AmSafe issued Service Letter (SL) No. 7041, Issue 1, for Cirrus Aircraft inflatable restraint assemblies. AmSafe noted customers reported an in-service problem of incorrect or missing anti-rattle plugs found on the belt buckle assemblies. AmSafe SL No. 7041 provided the procedure for removing and correctly installing replacement buckle/connector anti-rattle plugs. In response to AmSafe SL No. 7041, Cirrus Aircraft issued Service Advisory SA23-06R1 on April 25, 2023, to notify its customers of the issue and describe how to conduct a visual inspection of the buckles.

The right crew seat buckle anti-rattle plugs were found in the incorrect position. The buckle tang had its respective plugs in the correct position. The belt webbing was almost completely pulled through and was twisted around the load bar. The belt webbing did not exhibit any evidence of rubbing, discoloration or fraying. The buckle functioned normally when tested by hand, with no anomalies or damage noted.

The left crew seat’s restraint anti-rattle plugs were found in the correct position. The buckle functioned and no anomalies or damage were noted. The rear center seat restraint push-button buckle assembly did not exhibit any anomalies or damage, and the buckle functioned normally when tested by hand.

Airframe Parachute

The airplane was equipped with the CAPS, a whole plane ballistic parachute recovery system. The parachute canopy did not inflate during deployment and was found stretched-out on a linear trajectory into the surrounding treetops, consistent with the rocket motor deploying upon ground impact. The parachute enclosure cover was found near the main wreckage, also consistent with a CAPS deployment upon ground impact. TESTS AND RESEARCHFlight Testing

The National Transportation Safety Board (NTSB) Investigator-In-Charge (IIC) asked the airplane manufacturer, Cirrus Aircraft, to conduct several flight tests to determine if the recovered data from the accident airplane could be used to identify the root cause of the total loss of fuel flow during the flight.

The flight testing revealed that when the mixture control was moved full aft to the idle cutoff position with the electric fuel pump off, the decrease in fuel flow closely matched the recorded data from the accident flight, as shown in Figure 6. The fuel flow decreased from 7.6 gph to zero in 8-9 seconds. During the flight test, the throttle initially remained unmoved when the mixture control was placed in the idle cutoff position, but after a small rise in manifold pressure, the test pilot made one throttle reduction followed by two throttle advances to match the recorded accident data. Normal fuel flow and engine operation were restored when the mixture control was increased forward of the idle cutoff position.


Figure 6 – A plot comparing accident data with flight test data. The flight test scenario was the placement of the mixture control to the idle cutoff position with the electric fuel pump off. The plotted parameters include engine speed, fuel flow, and manifold pressure. The flight test data is time-aligned with the accident data to depict the loss of fuel flow at the same time.

Another flight test revealed that when the mixture control was moved full aft to the idle cutoff position with the electric fuel pump on, the decrease in fuel flow did not match the recorded data from the accident flight, as shown in Figure 7. Notably, the fuel flow did not decrease to zero; the fuel flow was 2.5 gph to 5.0 gph, depending on the throttle position. The test pilot noted that the engine continued to operate with noticeable roughness while the mixture control was in the idle cutoff position with the electric fuel pump on. Normal fuel flow and engine operation were restored when the mixture control was increased forward of the cutoff position.


Figure 7 – A plot comparing accident data with flight test data. The flight test scenario was the placement of the mixture control to the idle cutoff position with the electric fuel pump on. The plotted parameters include engine speed, fuel flow, and manifold pressure. The flight test data is time-aligned with the accident data to depict the loss of fuel flow at the same time.

In another flight test, a Cirrus Aircraft SR22 test airplane, operated in the experimental category, was modified with an electronic fuel valve installed before the inlet fitting to the electric fuel pump to simulate a separation failure of the 90° elbow inlet fitting. The electronic fuel valve, when actuated, resourced the electric fuel pump inlet from the fuel line to ambient air. The modified installation required an additional 6-inch segment of fuel line between the electronic valve and the electric fuel pump.

The modified test airplane was flown to altitude before the test pilot entered a descent with engine power settings similar to the accident flight. During the descent with the electric fuel pump off, the electronic fuel valve was activated to simultaneously shut off access to the fuel line and vent the electric fuel pump inlet to ambient air. The results of two similar tests did not replicate the total loss of fuel flow recorded during the accident flight, as shown in Figure 8. Notably, the fuel flow did not immediately decrease to zero, rather there was a slower decrease of fuel flow that extended past 25 seconds. In one test, the fuel flow eventually decreased to 0.02 gph about 30 seconds after the electronic fuel valve was activated. Normal fuel flow and engine operation were restored when the test pilot selected the electronic fuel valve to draw fuel from the fuel line instead of being vented to ambient pressure.


Figure 8 – A plot comparing accident data with flight test data. The flight test scenario was a simulated separation failure of the 90° elbow inlet fitting to the electric fuel pump with the electric fuel pump off. The plotted parameters include engine speed, fuel flow, and manifold pressure. The flight test data is time-aligned with the accident data to depict the loss of fuel flow at the same time.

Further flight testing was conducted to determine what affect an operating electric fuel pump had during a simulated separation failure of the electric fuel pump inlet fitting. The flight test data showed that the fuel flow did not immediately decrease to zero, rather an initial 20 second decrease toward a low fuel flow rate (< 0.5 gph) that continued another approximately 20 seconds before the test pilot restored normal fuel flow and engine operation by selecting the electric fuel valve to draw fuel from the fuel line instead of being vented to ambient pressure, as shown in Figure 9.


Figure 9 – A plot comparing accident data with flight test data. The flight test scenario was a simulated separation failure of the 90° elbow inlet fitting to the electric fuel pump with the electric fuel pump on. The plotted parameters include engine speed, fuel flow, and manifold pressure. The flight test data is time-aligned with the accident data to depict the loss of fuel flow at the same time.

Materials Laboratory Examination

The separated 90° elbow fitting and adjacent flexible fuel line were examined by the NTSB Materials Laboratory. According to documentation from the manufacturer, the fitting designation was MS51527-B8ZC, a nominal 0.5-inch-diameter fitting made from plain carbon steel with a zinc-plated finish. One end was a boss-type connection, with straight threads, a flat end, and an O-ring, backing washer, and compression nut for mounting to the inlet of the electric fuel pump. The other end was a straight-threaded 37° flared tube for attaching a compression-type B-nut fitting. The flexible fuel line assembly, terminated by its own transition elbow with B-nut connection was attached to the flared tube end of the fitting.

The fitting separated at a copper-brazed joint within the fitting, adjacent to the 90° bend. The joint consisted of a round socket machined into the block that contained the bend and a tube that was pressed into the socket, with braze metal joining the two sides. When the fitting separated, the socket-side stayed attached to the electric fuel pump and the flared tube-side stayed attached to the flexible fuel line.

Both sides of the joint were plastically deformed and the B-nut connection between the fuel line and the fitting showed signs of counterclockwise rotation of the fuel line’s transition elbow relative to the B-nut and B-nut relative to the brazed fitting.

On one side, the tube was compressed in the axial direction, deformed radially outward at the butt-end of the tube, and deformed radially inward near the inlet to the joint. On the opposite side, the tube was deformed radially inward at the butt-end. A stripe of adhesive (torque stripe) that had been applied to the brazed fitting, B-nut, and transition elbow was cracked at the two transitions (from brazed fitting to B-nut and B-nut to transition elbow) and offset in the counterclockwise direction. The features were consistent with a load acting on the joint that caused the tube to lever out of the socket and a moment acting on the transition elbow in the counterclockwise direction.

A primarily silicon residual film was observed on external zinc-plated surfaces adjacent to brazed joint on both the tube-side and socket-side of the fitting. The film on the tube side was clear. However, parts of the film on the socket-side were clear while others had a blue tint that was consistent with absorbed fuel dye.

Simulator Sessions

A NTSB Human Performance Specialist observed two simulator sessions to evaluate a normal RNAV runway 17R approach and then to simulate and document a fuel-related loss of engine power emergency. The simulator used was based on a Cirrus SR20, not an SR22 or SR22T. The simulator was equipped with the Cirrus Perspective+ integrated avionics system that was consistent with the accident airplane, but it was equipped with 10-inch displays rather than 12-inch displays.

The NTSB Human Performance Specialist first documented the primary flight display (PFD) with descent profile during the RNAV approach, the fuel flow indicator on the multi-function display (MFD) and captured the audio call out for 600 ft agl as a method of indicating when the airplane was in the desirable parameters for CAPS deployment.

The second simulator session documented the visual and auditory cues of a fuel emergency was taking place. The crew-alerting system (CAS) included white, yellow, and red visual warnings displayed on the PFD. The main MFD page and the engine MFD page, displayed various engine parameters including, but not limited to, fuel flow, manifold pressure, and engine RPM. The auditory indication of a loss of engine power was a decrease in engine noise. There are no discreet aural warnings for a loss of fuel flow pertaining to the fuel mixture control lever being in the cutoff position.

Inadvertent Movement of Fuel Mixture Control

During one of the simulator sessions, the NTSB IIC inadvertently moved the fuel mixture control lever aft while cycling through power (throttle) positions while he was seated in the left crew seat of the simulator. Figure 10 shows an exemplar power (throttle) lever, mixture control, and the three-position electric fuel pump switch for the Cirrus SR22. It was determined that the NTSB IIC’s inadvertent movement of the mixture control was due to his hand/finger placement on the throttle’s T-handle, and the throttle had to be moved to the full idle position to inadvertently move the mixture control to the cutoff position. The Flight Training Development (FTD) Specialist serving as the simulator technician remarked that it was not something he had ever seen happen before. Additionally, the FTD specialist discussed and demonstrated alternative hand/finger positions that he typically uses to move the throttle that did not interfere with the mixture control.


Figure10 – Power (throttle) lever, mixture control, and electric fuel pump switch

When asked during interviews, none of the factory flight instructors or supervisors had witnessed a fuel mixture control lever inadvertently pulled completely to cutoff; however, one instructor stated that he witnessed the cord of a headset get wrapped around the fuel mixture control knob, which leaned out the mixture to the point that the engine started running rough. It was the rough running engine that alerted the instructor to the problem. Another flight instructor said that he had seen a headset cord close enough to the throttle and fuel mixture control lever to make him uncomfortable enough to move it, but that it did not directly interfere with the engine controls.