N36SGAgusta Aerospace AW139 · 2010

On April 6, 2013, about 1035 Pacific daylight time, a Beechcraft A36TC airplane, N36SG, was substantially damaged during a forced landing in Big Bear Lake, Big Bear City, California. The pilot received serious injuries, and his three passengers received minor injuries. The personal flight was conducted under the provisions of Title 14 Code of Federal Regulations Part 91. Visual meteorological conditions prevailed at the accident location. The airplane had departed on an instrument flight rules (IFR) flight plan from McClellan-Palomar Airport (CRQ), Carlsbad, California, and was destined for Big Bear City Airport (L35).

According to the pilot, his passengers include his niece, her husband, and their 5 year-old son. En route, the pilot canceled his IFR flight plan, and continued, using flight following services from SoCal Approach. The pilot planned to follow his normal route, which included a slow letdown to L35 after passing ARRAN intersection. Shortly after the airplane passed ARRAN, at an estimated altitude of between 9,500 and 9,000 feet, the airplane experienced a "very sudden" onset, high-frequency vibration. The pilot reduced the power setting, and notified the air traffic controller of the problem. Initially the manifold pressure gauge provided very erratic indications, but the pilot was unable to discern whether this was a valid indication, or the instrument responding to the vibration.

The airplane began to descend more rapidly than the pilot desired, and he adjusted the engine controls to yield normal cruise manifold pressure and rpm values. However, the airplane did not respond commensurately with the indicated power setting, and continued its descent. When the airplane was approximately abeam the "observatory" (a visual landmark) at an altitude of about 6,900 feet, the pilot switched to the airport advisory frequency, and notified them of his problem. The pilot observed a causeway that crossed the lake between his position and the airport, and began a turn to his right (south) to avoid the causeway. At some point, the pilot recognized and broadcast that he was unable to make the airport, and planned to land in a field west of the airport. The pilot banked to the right to turn for the field, but the right wingtip contacted the lake, which ended the flight. The airplane came to rest inverted in water about 2 feet deep. The impact location was about 0.3 miles southwest of the causeway, and about 1.25 miles southwest of the L35 runway 8 threshold.

An eyewitness who was located on his property at the lakeshore was attracted to the airplane by its sound. He estimated that at that time he noticed the airplane, it was about 30 feet above the lake, and headed south, towards him. He then saw it bank right, and impact the lake. He summoned 911 assistance, and then ran into the lake to render assistance.

According to Federal Aviation Administration (FAA) records, the pilot held a private pilot certificate with airplane single engine land and instrument airplane ratings. He reported that he had a total flight experience of approximately 1,300 hours, including about 1,250 hours in the accident airplane make and model. His most recent flight review was completed in April 2012, and his most recent FAA third-class medical certificate was also issued in April 2012.

FAA records indicate that the airplane was manufactured in 1979. It was equipped with a Teledyne Continental TSIO-520 series engine, and a McCauley model 3A32C76 propeller. The airplane was originally registered as N66656. In 1986 the registration was changed to N36ML, and in 1996 to N36SG. The pilot purchased the airplane in November 1998.

Maintenance records indicated that the airplane had been upgraded with multiple avionics changeouts or additions, as well as aftermarket wing tip tanks. Both the accident engine and the accident propeller hub were different serial numbers than the ones that were originally installed during manufacture of the airplane. However, all three accident propeller blades were the same ones that were originally installed during manufacture of the airplane.

The L35 1035 automated weather observation included winds from 260 degrees at 8 knots, variable between 224 and 284 degrees, visibility 10 miles, clear skies, temperature 15 degrees C, dew point minus 8 degrees C, and an altimeter setting of 30.14 inches of mercury.

FAA records indicate that the airplane was manufactured in 1979. It was equipped with a Teledyne Continental TSIO-520 series engine, and a McCauley model 3A32C76 propeller. The airplane was originally registered as N66656. In 1986 the registration was changed to N36ML, and in 1996 to N36SG. The pilot purchased the airplane in November 1998.

Maintenance records indicated that the airplane had been upgraded with multiple avionics changeouts or additions, as well as aftermarket wing tip tanks. Both the accident engine and the accident propeller hub were different serial numbers than the ones that were originally installed during manufacture of the airplane. However, all three accident propeller blades were the same ones that were originally installed during manufacture of the airplane.

The airplane came to rest inverted in the lake, in about 2 feet of water, and about 100 feet from shore. The impact location was approximately 1.25 miles southwest of the L35 runway 08 threshold. The left wing had separated from the fuselage during the accident sequence, and was located about 75 feet from the airplane. Aside from loss of several transparencies, the cabin remained essentially intact, and did not lose any appreciable occupiable volume. The forward and aft cabin doors remained operable, and the pilot and passengers exited through those doors.

The airplane was recovered by dragging it inverted onto shore, where it was then righted. Some fuselage deformation resulted during the lake-recovery efforts. The airplane was transported to a secure facility for additional examination. Detailed examination of the engine and airframe did not reveal any pre-impact conditions or anomalies that would have precluded normal operation and continued flight.

The three blade, variable pitch propeller remained attached to the engine propeller flange. All three of the propeller blades were retained in the propeller hub; however, all three propeller blades were free to rotate independently of one another in the hub. All three of the propeller blades displayed aft bending, and two of the propeller blades displayed twisting deformation. The propeller flange remained attached to the rest of the crankshaft and there were no cracks noted in the crankshaft. The propeller was removed from the engine and shipped to McCauley, the propeller manufacturer, for detailed examination.

The airplane was equipped with a JPI EDM-700 engine monitor, which displayed multiple engine parameters in the cockpit, and stored time history data of those parameters in non-volatile memory. The unit was removed and sent to the NTSB Recorders laboratory in Washington DC for data download. The accident flight was successfully downloaded, but review of the recorded information did not reveal any engine-related anomalies.

The airplane was equipped with a Shadin 1530 fuel flow monitor, which displayed fuel-related information in the cockpit, and stored time history data of those parameters in non-volatile memory. The unit was removed and sent to the NTSB Recorders laboratory in Washington DC for data download. The accident flight was successfully downloaded, but review of the recorded information did not reveal any fuel-related anomalies.

Propeller Design

The propeller was a constant-speed design which permitted the blades to change pitch in flight to maintain a preset rpm selected by the pilot. Pitch change commands were effected by pressurized engine oil, metered by the propeller speed governor, that drove a piston fore and aft inside the hub, which, through a series of links, converted the piston translation into blade rotation.

Each propeller blade had ring-shaped steel ferrule attached to its blade butt. The ferrule was oriented so that its circular plane was perpendicular to the span of the blade. Once installed, the ferrule did not move relative to the blade. Each ferrule measured about 5 inches in outside diameter, and about 1 inch thick. The face that was oriented towards the propeller axis of rotation (center of the hub, away from the propeller blade) was referred to as the "forward" face; this is not a reference to its orientation in the airplane axis system.

Each propeller blade was also equipped with a steel actuating pin. Each pin measured about 3 1/2 inches in length. The "upper" two-thirds of the overall pin length consisted of a constant- diameter rod that terminated in a smooth cone at the top. The lower one-quarter of the overall pin was threaded, and a hexagonal set of wrenching flats was situated between the threaded section and the upper smooth rod. The wrenching flat section was approximately twice the diameter of the other two sections, and each of those sections was smoothly faired into the wrenching flat.

An actuating pin was threaded into the forward face of each ferrule. The pin extended parallel to, but in the opposite direction of, the blade. A special washer served as part of the interface between the pin and the ferrule. One side of the washer was planar with concentric grooves. The center hole was chamfered on the opposite side to accommodate the stress relief fillet of the pin. The planar side of the washer abutted the ferrule face, and the chamfered side of the washer accommodated the stress relief fillet of the pin that was situated between the pin threads and the wrenching flats. When properly installed, there was no relative motion between the pin, washer, and ferrule. The wrenching flats and the ferrule had accommodations for safety-wiring the pin to the ferrule to prevent the pin from backing out of the ferrule.

The actuating pins were primary components of the propeller blade pitch control linkage. They converted the linear piston motion into rotary motion of the propeller blades in the hub, and also served to hold the blades at the commanded pitch setting. Due to the combination of operating airloads on the propeller blades, and the positioning loads of the propeller hub piston, the actuating pins were continuously loaded in bending. Load magnitudes and directions were not constant. The pin, washer, and ferrule design was such that, when properly installed, the bending loads were resisted by the ferrule, washer, and the portion of the pin "above" (protruding from) the ferrule face; no bending loads were transmitted to the (primarily threaded) portion of the pin "below" the ferrule face. Failure to properly seat the pin in the washer and seat the washer against the ferrule face would permit bending loads to be transferred to any portion of the pin not engaged in the ferrule threads.

Ferrule and Actuating Pin Design Changes

The original design of the ferrules utilized through-holes for the actuating pin installations; the threaded hole for the pin extended completely through the ferrule. In 1972, due to ferrule failures at the through-holes, McCauley issued Service Bulletin (SB) 99. SB99 was considered "necessary and required" by McCauley, and was to be accomplished at the next propeller overhaul.

SB99 mandated the use of modified or new (redesigned) ferrules and actuating pins across a wide range of propeller models, including the accident propeller model. The SB eliminated the use of ferrule through-holes, and instead mandated the use of blind holes that only penetrated one surface of the ferrule. In conjunction with the shallower blind hole, the threaded section of the actuating pin was reduced by approximately 0.085 inches.

For the accident propeller model, the superseded (through-hole) ferrule part number (PN) was C-3111, and was replaced by the blind-hole ferrule PN C-4451. In addition to a lack of through-holes, new PN ferrules bore a machined circumferential groove on their outer diameter. Superseded ferrules with through-holes could be modified to the new blind hole configuration by disabling or removing the threads in the through hole, and drilling and threading a new blind hole.

The ferrule face opposite the forward face was referred to as the "aft" face. SB99 explicitly warned against the creation of bulges on the aft face as a result of the drilling, and prohibited the use of any ferrules that incurred or exhibited such bulging.

The superseded (longer) actuating pin PN was B-3491, and the new (shorter) actuating pin PN was B-4460. In addition, although manufactured of the same material, the required hardness of the new actuating pin was increased. SB99 permitted the use of superseded ferrules with new blind holes, and the use of superseded actuating pins with threaded portions reduced in length by cutting.

Once the new pin design was implemented, the superseded pins were no longer produced; the new and old pins were not produced concurrently. The engineering drawings for both the superseded and new PN actuating pins did not require the pins to be permanently marked with part numbers. The SB required that superseded actuating pins modified (shortened) for re-use were to be identified by a stamped "XR" on the bottom of the threaded end.

SB99 contained an explicit warning that prohibited the installation of an unmodified, superseded (long) actuating pin in a blind ferrule hole.

Propeller Maintenance Practices

Propeller overhaul requires the separation of the steel ferrule from the aluminum blade in order to permit all necessary component inspections. If the ferrules and blades are in airworthy condition and will be reused on the same propeller, the ferrule-blade pairing will be altered; each ferrule is assigned to (and eventually installed on) a blade different than the one from which it was removed. This re-pairing is accomplished because reinstallation on same blade will typically not provide for accurate blade pitch range when the same pin hole is utilized, or will result in insufficient edge-distances for newly-drilled pin holes.

There was no requirement to retain or reuse the same blades or ferrules on the same propeller at overhaul; those components were not matched sets and could be sourced as needed by the overhauler.

The maximum permitted number of ferrule re-uses (and thus newly drilled pin holes) was three, which would result in a maximum of four holes in a single ferrule.

Since at least December 1979, with the original issue of SB137, the McCauley-specified overhaul limits for the accident propeller model were every 60 calendar months (from engine installation) and/or 1,500 hours in service, whichever came first.

Propeller Examinations

Initial examination of the propeller was conducted at the facilities of McCauley Propeller and Cessna Aircraft. Disassembly of the hub revealed that it contained multiple fractured components, sandy/gritty material, and sludgy oil.

The hub serial number indicated a 1991 date of manufacture. The blade serial numbers were close together, and indicated a much older date of manufacture, prior to 1970. The propeller had been field modified to oil-filled from the original grease-filled configuration. The propeller had indications of prior dynamic balance. Clean reflective tape, a decal, and balance weights were observed installed on the spinner bulkhead.

The three actuating pins were found loose inside the hub, separated from their respective blades. One actuating pin had fracture-separated from its base, which remained threaded into its ferrule. The other two actuating pins, although bent and otherwise damaged, retained their full length. The threads of those two pins exhibited significant damage, consistent with the pins being forcibly extracted ("stripped"), instead of unscrewed, from the ferrules. The damage to the actuating pins precluded determination of the lengths of their threaded portions. The actuating pins or ferrules from all three blades contained remnants of safety wire used to secure each pin to its respective ferrule. None of the pins bore any part numbers, stamps, or any other identification markings.

Visual and microscopic examination revealed that the fractured actuating pin had failed in fatigue. The fatigue failure origination point was located at the thread root, where the threaded portion of the pin begins. The crack initiation site did not contain any indications of material deficiencies such as inclusions, pits, or voids. The two intact actuating pins exhibited some cracking on the tension side of their pin thread roots. Those cracks were consistent with overstress.

Independent tests at the NTSB and Cessna materials laboratories yielded similar results regarding the actuating pins. The hardness values for all the actuating pins were lower than the current engineering drawing requirements; they were in line with the specified hardness values for the superseded pins. The measured hardness values ranged from HRC (Hardness, Rockwell C scale) 29 to 31.4, compared to the design-required range of HRC 34 to 38. Chemical composition for the two bent/damaged actuating pins was in accordance with 4140 steel per the engineering drawing. NTSB chemical composition testing of the fractured pin was conducted using energy dispersive spectroscopy. The results were consistent with the design materials, AISI 4130 or 4140 steel. The Cessna laboratory did not have sufficient fractured pin material to enable a chemical composition analysis.

The three ferrules all remained attached to their respective blades. The ferrule with the base remnant of the fractured pin still installed had a total of two drilled pin holes, both of which were blind holes. The other two ferrules, which did not contain their actuating pins, each had a single, blind, pin hole. The observed mismatch in the number of actuating pin holes in the ferrules was atypical, but not prohibited.

Sections through the broken actuating pin and the surrounding ferrule revealed that the actuating pin contacted the bottom of its blind hole ('bottomed out') in the ferrule. Because the two damaged but unbroken pins were found separated from their respective ferrules, the investigation was unable to determine whether or not those two pins were bottomed out in their blind ferrule holes.

Examination of the ferrule that contained the fractured pin remnant revealed that the aft surface was bulged outward in a location 'beneath' the blind hole where the broken pin section remained. In addition, a wear pattern was present on the bearing race, consistent with contact between the bulged ferrule surface and the bearing race. Examination of the aft surface below the adjacent, but empty, blind hole revealed that that ferrule did not exhibit any deformation or bulging in that location. The aft surfaces of the ferrule flanges from the other two blade assemblies also exhibited bulging deformation below their blind holes.

The microstructure for all three ferrules was typical for heat treated 4140 steel, per design. The hole depths on all three ferrules measured in the range of 0.349 to 0.355 inches; the engineering drawing depth specification was 0.340 to 0.350 inches.

The forward face of the ferrule with the fractured actuating pin section exhibited imprints from the washer that was installed between the pin and the ferrule. The imprint was uneven in depth, consistent with the plane of the washer bottom face being parallel to the plane of the ferrule face. Dimensional checks of the washers revealed that all were in accordance with the design specifications.

Propeller examination report is contained in the public docket for this accident.

Propeller History and Maintenance Records

Review of all provided maintenance records revealed two propeller overhauls, one in 1992, and another in 1998. No records prior to 1992 were provided.

The 1992 overhaul was accomplished by American Propeller Service in Redding California. Those records indicated that the airplane tachometer time was 296.2 hours. An annotation that appeared similar to "TT" was accompanied by an entry of 1980.2. The investigation was unable to determine the meaning of this entry.

The propeller blade serial numbers in those 1992 records were the same as the serial numbers of the blades on the airplane at the time of the accident. The hub serial number listed in the 1992 records did not match the serial number of the hub on the airplane at the time of the accident. The records did not contain any other significant details regarding the overhaul.

The most recent propeller overhaul was accomplished by Aircraft Accessories of Tulsa Oklahoma in February 1998. The serial numbers of the hub and propeller blades in those 1998 records were the same as the serial numbers of the hub and propeller blades on the airplane at the time of the accident. The maintenance entry stated that "all pertinent" manufacturer's service bulletins through SB218 had been complied with.

Review of the parts lists from the 1998 overhaul records did not indicate whether the hub or actuating pins were replaced with new components, the previous serviceable components, or other serviceable components. The overhaul records did not contain any references to the actuating pins, including any references to the actuating pin part numbers. The parts list contained two lines regarding ferrules. The first line listed "C4451 ferrules" in the "PART # & NAME" column, and a "2" in the ""QTY" column. The line just below that contained two double quotation/ditto marks followed by "(seviceable)" [serviceable] in the "PART # & NAME" column, and a "1" in the "QTY" column.

Review of the maintenance records revealed that the airplane had been continuously maintained in an airworthy condition, by means of annual inspections, and scheduled and unscheduled maintenance, since at least 1998.

The most recent airplane annual inspection, and propeller 100-hour inspection, was completed in October 2011. At the time of the annual and 100-hour inspection, the propeller had a total time in service of 4,828.46 hours, and a time since overhaul of 1,310.53 hours.

At the time of the accident, the propeller had accumulated approximately 15.2 years and 1,361 hours in service since its most recent overhaul. Although the propeller had not exceeded the propeller manufacturer's overhaul service limit of 1,500 hours, it exceeded the manufacturer's calendar time limit of 60 months by a factor of three.

While en route the pilot canceled his IFR flight plan, but continued to communicate with SoCal Approach for flight following services. All communications, including those during the IFR portion of the flight, were normal until about 1033. At that time the pilot informed the air traffic controller that he had a "serious vibration," but that he was still planning to continue to his destination, L35. The controller asked the pilot whether he knew what the cause of the vibration was. The pilot responded in the negative, and described some of his instrument indications. This was followed by a brief exchange about the number of persons and amount of fuel on board. About 1035, the controller asked the pilot whether he was able to maintain his altitude, and the pilot responded that he believed he could. The controller then advised the pilot to hold his altitude as long as possible, and to "circle down" over the airport once he reached it. The pilot concurred with that strategy.

The controller then provided a telephone number and requested that the pilot call him to advise that he had landed safely, which the pilot agreed to do. About 1036, when the airplane was about 7 miles southwest of the airport, the pilot queried the controller as to when he thought he should switch to the L35 "CTAF" (common traffic advisory frequency). The controller provided the traffic that he observed to the pilot, and then instructed the pilot to contact CTAF. The pilot agreed, and no further communications occurred between the pilot and SoCal Approach.

The L35 CTAF communications were not recorded, but the radio operator on the ground at L35 provided a written statement of his recollections. According to the radio operator's statement, the pilot contacted CTAF and advised that his airplane had a serious vibration problem. The radio operator "cleared" him to land on runway 8, and also broadcast an announcement for all other aircraft to "stay clear" to allow the subject airplane to land. The pilot then announced that he was unable to make the airport, and was planning to land in a field. No other communications were received from the airplane.