N135WTAirplane Factory SLING 4 TSI · 2018

On September 25, 2023, about 1250 Pacific daylight time, an experimental amateur-built Sling 4 TSI, N135WT, was substantially damaged when it was involved in an accident near San Pedro, California. The pilot and passenger were seriously injured. The airplane was operated as a Title 14 Code of Federal Regulations Part 91 instructional flight.

According to FAA ADS-B data, the airplane departed about 1223. After a short discovery flight near the coastline, the instructor turned to a northerly heading to return to the airport. About 1246, the airplane began a descent from an altitude of about 2,400 ft mean sea level. At 1246:46, the pilot reported to air traffic control that he had “an engine failure” and requested approval to land. The controller immediately cleared the airplane to land and the pilot acknowledged the clearance. He then asked the pilot for his desired runway and the pilot responded with the “south runways,” referring to runways 29R and 29L. There were no further transmissions from the pilot.

A witness video recorded the airplane at about 1249 as it flew north trailing white smoke. About 12 seconds later the smoke ceased and the airplane subsequently began a right descending turn. The airplane then entered a steep right turn in a nose-down attitude and rapidly descended towards the ground as it disappeared.

The airplane impacted a field about 3 nm southeast from the pilot’s destination airport, Zamperini Field Airport (TOA), Torrance, California.

Engine data ECU

The engine was equipped with an engine control unit (ECU) that recorded engine telemetry. According to the ECU data, the airplane departed on the accident flight with 687.90 engine hours and captured the first faults at 688.50 engine hours (716:16:19 ECU hours), indicating that the engine had been operating for about 36 minutes when the loss of power occurred.

The linearized throttle position was consistent with the engine speed for most of the accident flight, consistent with the engine responding to the pilot's inputs. At 716:08:20 ECU time, the pilot set the throttle to 94%, with the engine running at 5,428 rpm. Three minutes and 20 seconds later the engine power began to fluctuate before it started to lose power. The pilot advanced the throttle to full power (100%) and then cycled the throttle between 25% and 100% multiple times; however, the rpm decreased to about 2,615 rpm and then advanced to about 4,300 rpm with the throttle position at 100%. At 716:13:10, the engine speed briefly recovered to 5,303 rpm and then made a gradual decline until it lost all power about 3 minutes later at 716:11:40.

The exhaust gas temperatures, fuel injection mass and air/fuel ratio, and knock events and ignition timing data all exhibited abnormal values after the engine operation became erratic. The exhaust gas temperature (EGT) for cylinder No. 4 showed a gradual rise from about 785°C to a peak of 898°C by the time the engine power failure began at 716:11:39. Three seconds later the No. 4 cylinder EGT dropped to 557°C and continued to decrease to 331°C by the time the engine lost all power. According to the engine manufacturer, the fuel injection mass was consistent across all cylinders and the metered air/fuel ratio targets were consistent with a relatively lean air/fuel mixture before the engine failure at 716:11:39. When the failure began, some of the fuel injection mass data points were 0 by 716:11:57. The metered air/fuel ratio target values decreased below 13.80 by 716:11:41. The ECU recorded a total of 350 knock (detonation) events from 712:30:40 through 716:11:39 for cylinder No. 4 and 87 total knock events for cylinder No. 3. In addition, the ECU is designed to adjust ignition timing to mitigate knock events. Before the failure at 716:08:20, the ECU captured isolated knock events that resulted in a timing retard of -0.5° after top dead center (ATDC) on the No. 4 cylinder. After the failure, the ignition timing retarded up to -6.27° ATDC and noted the correction “Timing Retarded,” reflecting the ECU’s effort to protect against potential damage due to uneven combustion.

According to the engine manufacturer, persistent knock activity suggests an ongoing or worsening condition. The ECU data is consistent with the No. 4 cylinder exhibiting constant stress leading up to its failure, marked by elevated EGTs, fuel irregularities, and chronic knock events. After the failure at 716:08:20, the engine speed became unstable, which was accompanied by manifold air pressure fluctuations and increased knock mitigation efforts.

Garmin G1000

According to data retrieved from the onboard multi-function display of the Garmin G1000, the exhaust gas temperatures (EGT), engine speed, manifold pressure and fuel flow were all consistent for the entire accident flight. Although the engine data was consistent with the ECU data, the two could not be correlated. The data showed an unremarkable takeoff and climb. At 12:23:30, EGTs rose to about 1,700° F, accompanied by a rise in engine speed to about 5,500 rpm with corresponding increases in manifold pressure and fuel flow.

At 12:35:45 the EGTs, engine speed, fuel flow, and manifold pressure reduced along with a sudden reduction in oil temperature and increase in oil pressure for about 2 minutes and 15 seconds and then returned to their original values. At this time the GPS altitude decreased from about 7,000 ft to 0 over the next 14 minutes (until the end of the flight) with a brief rise at 12:46. Engine performance remained steady until 12:45:50 when the EGTs, engine speed, fuel flow and manifold pressure reduced again. The EGTs for the Nos. 2 and 4 cylinders points ceased at about 12:46:50; however, fuel flow, manifold pressure, oil pressure, and the EGTs for cylinder Nos. 1 and 3 increased and became erratic for the remainder of the flight. The engine power decreased again at 12:50:18 and the last recorded data was captured at 12:50:32.

Postaccident examination revealed no preaccident mechanical anomalies or malfunctions with the flight control system that would have precluded normal operation.

The postaccident engine examination showed that the No. 4 cylinder spark plugs were mechanically damaged and the cylinder’s exhaust and intake valves had separated. There was also a hole in the piston and mechanical damage throughout the piston face.

An engine teardown revealed that both valves on the No. 4 cylinder had separated at their valve stems. The intake valve was found inside the cylinder and the exhaust valve was found inside the airbox, both with extensive mechanical damage, gouges, and scoring. The upper portion of the valve stem below the keeper exhibited a hardened black material with discoloration along the shaft of the valve stem.

The Nos. 1 and 3 cylinder piston faces and valves were unremarkable and exhibited coloration consistent with normal wear. Cylinder No. 2 exhibited gouge marks along the circumference of the piston and near the center. The No. 2 cylinder intake and exhaust valves displayed some pitting and gouge marks, and metal fragments were embedded into the cylinder head. Disassembly of the cylinder valve assemblies did not reveal any mechanical anomalies with Nos. 1, 2, and 3 cylinder rocker assembly, coils, keepers, washers, or push rods.

An NTSB materials laboratory examination of the cylinder assemblies found that the black residue on the valve stem of the exhaust valve was comprised of lead and carbon, consistent with leaded fuel combustion byproduct. The residue prevented the removal of the valve stem from the valve guide until the residue was removed. Brown and black discoloration was observed on the stems of each exhaust valve with the most extensive coloration on the No. 4 cylinder exhaust valve. The No. 4 cylinder exhaust valve was mechanically damaged.

According to the heavy maintenance manual, oil residues are permissible on the valve stem up to 20% of the running surface or the valve may be at risk of failure at the weld point due to overheating. The manual further states that any valve showing higher amounts of residue must be replaced due to increased risk of failure at a weld in the valve. The No. 4 cylinder exhaust valve discoloration covered 60% of the running surface of the valve stem.

The heavy maintenance manual also included tolerance limits for the valve guides and the minimum and maximum tolerance for the space between valve guide and valve stem (or backlash). Cylinder No. 4 was at or above the 100% tolerance limit at all measured locations in the valve guide. Both the No. 3 and 4 cylinders were above the 100% tolerance limit for the maximum backlash, measured at the bottom of the running surface.

Figure 1: The four exhaust valves and their discoloration on the running surface of the valve stem

Valves

The engine was equipped with sodium-filled valves, which have a hollow cylindrical chamber in the stem partially filled with sodium metal. During engine operation, the sodium transfers excess heat from the valve head to the stem. Contact between the valve stem and valve guide transfers heat to the cylinder head via conduction. According to a metallurgical examination, if there is excessive clearance in the valve guide, the heat in the valve stem is not transferred to the cylinder head, and the valve will overheat.

Rotax Conclusion of Findings

According to the engine manufacturer, excessive valve wear can result from overheating, combustion deviation, insufficient lubrication and also noted the following:

· Excessive carbon buildup on the valve neck and adjacent stem can lead to stem scuffing and excessive wear.

· Insufficient lubrication or breakdown of valve stem lubrication can cause scuffing and accelerated wear.

· Abrasive particles contaminating the valve stem and guide bore can result in accelerated or excessive wear.

· Prolonged operation with AVGAS 100LL combined with infrequent oil change intervals can accelerate valve and guide wear and oil sludge buildup.

This information was provided by the engine manufacturer after the NTSB metallurgical examination was completed. The NTSB database did not contain any records of valve failures from Rotax 915 model engines that were the result of the prolonged operation of 100LL aviation grade gasoline.

The airplane operator noted that they operate their airplane fleet with a mix of 91 octane MOGAS and at least 30% 100 low lead aviation grade gasoline.

Section 05-10-00 (page 11) of the engine line maintenance manual stated the following points of inspection should take place at the corresponding intervals:

Table 1: Oil inspection/service table (*required in the case more than 30% of operating with leaded fuel such as AVGAS)

The engine manufacturer’s line maintenance manual also did not contain any reference to the effects of the prolonged exposure of 100LL aviation grade gasoline.

The engine logbook showed that none of the required oil inspections or maintenance were performed at 50-hr intervals as required by the maintenance manual. According to the operator, their maintenance team was under the impression that their blend was consistently 23% AVGAS, which did not require them to perform the 50-hr oil system maintenance inspections. However, after the accident the operator discovered that their technique for measuring this ratio was inconsistent and produced ratios that included anywhere between 25% and 35% of AVGAS.

Leaded Fuel

The manufacturer’s line maintenance manual also contained numerous warnings concerning the use of leaded fuels (e.g. AVGAS). Section 12-20-00 (page 46) warns against increased wear of the spark plugs. The manual also notes that the 50-hr check is only required with the use of leaded AVGAS.

“This maintenance schedule contains a column for a 50 hr. check. This check is recommended by the manufacturer but not essential, with the exception of oil change when operating with leaded AVGAS.”