Close-Up: The Jeff Ethell P-38 Crash

Accident occurred JUN-06-97 at TILLAMOOK, OR
Aircraft: Lockheed P-38L, registration: N7973
Injuries: 1 Fatal.

The aircraft (P-38) had been topped off with full reserve & main fuel tanks (44 USgal/engine reserve tanks & 72 gal/engine main tanks). The pilot took off with about 20minutes fuel having been already consumed. Operating with another P-38 within the airportarea, the 2 aircraft flew about 20 to 25 minutes each. The fuel consumption was reportednominally at 60 gallons/hour/engine (1 gal/minute). With 44 gallons of fuel in eachRESERVE tank for the duration of both flights, the engines would have exhausted allavailable fuel in each RESERVE tank after about 44 minutes. Both fuel selectors were foundon the RESERVE setting at the site. No mechanical malfunction was found with eitherpropeller or engine. The flaps & landing gear were retracted. According to the Pilot'sManual (flight manual), if one engine fails below 120 mph (safe single-engine airspeed),the pilot is to "close both throttles and land straight ahead." The flightmanual did not provide any information for aircraft minimum control speeds with the flapsfully retracted. Several witnesses reported the aircraft was slow while turning base.Since this was a single-seat aircraft, there was no provision for "dual"instructional training in single-engine procedures or spin recovery. The pilot wasreported to have flown 6 or 7 hours in another P-38, which included practice simulatedsingle-engine maneuvers, but no actual in-flight shut down and feathering of an engine.

History Of Flight

On June 6, 1997, approximately 1800 Pacific daylight time, a Lockheed P-38L"Lightning," retaining a limited airworthiness registration, N7973, registeredto Bruce L. Pruett, and being flown by a commercial pilot, was destroyed during collisionwith terrain following a loss of control while on visual approach to the Tillamookairport, Tillamook, Oregon. The pilot was fatally injured and a post crash fire destroyeda portion of the aircraft. Visual meteorological conditions existed, and no flight planhad been filed. The flight, which was a practice run for an upcoming P-38 pilot convention(including the pilot's father, a World War II P-38 pilot), was to have been operated under14CFR91.

The pilot/owner of N2114L, a P-38L sister ship nearly identical to N7973, and which wasrestored to flying condition several months earlier, was interviewed. He reported thatboth N2114L and N7973 took off from the Tillamook airport between approximately 1715-1730.He reported logging approximately 20-25 minutes of flight time on the flight (in N2114L).He also reported taking off immediately before N7973 and was ahead of the aircraft duringhis approach and landing.

An 8 millimeter video recording, made by the manager of the Tillamook Naval Air StationMuseum, and showing the takeoff of both P-38's, was reviewed. It was noted that bothaircraft started their engines nearly simultaneously and N2114L (the olive green P-38L)departed immediately ahead of N7973 (the silver P-38L). The duration of this tape, whichconcluded with N2114L taxiing in to the ramp following its flight, was measured on astopwatch and found to be 9 minutes and 57 seconds in duration. A number or recognizablebreaks in the video recording were noted as the operator, who was recording from the rampat the Tillamook airport, stopped recording and then began again at later times during theflight duration. N7973 was observed to make a number of low passes (with N2114L) over theairport, one of which included an aileron roll. Both aircraft remained within severalthousand feet above ground maximum elevation during their flights.

A number of witnesses observed the aircraft during the final moments of flight. Thefollowing observations were reported be these witnesses as described in their statements(attached):

Witness Simmons, who was located approximately one-quarter mile northwest of the crashsite in a vehicle travelling eastbound on Simmons Creek Road, reported seeing "awobbling, or wavering of the wings" and that the "last thing I saw was the P-38in a nose-down position."

Witness Johnston, who was located approximately one mile north of the crash site at hisresidence (looking south) reported seeing the P-38 "in gentle circles going towardsthe ground" and stated that he saw "at least four revolutions counterclockwise(left to right) before it went behind the hill." Additionally, he stated that "Ialso had the impression that the aircraft was in a flat spin along with the slowrevolutions to the earth."

Witness Tone, who was located approximately one mile south of the crash site at herresidence (looking north) reported seeing the plane "heading east just behind sometrees. It came out into a small clearing and I noticed how low it was flying."Additionally, she stated that "it was going very slow like it was gliding and then itstarted slowly flipping out of control, not end over end. It went more sideways (and)rolling over slowly."

Witness Nielsen, who was located approximately one mile southwest of the crash site ata residence (looking northeast) reported seeing the plane "going east at a goodheight flying almost level when (he) heard a pop-pop. The next time (he) looked up it wasin a tailspin."

Witness Larsen, who was located approximately one-half mile west of the crash site at aresidence (looking east) reported seeing the plane "flying east at an altitude ofapproximately 400-500 feet, when it began to sputter (and) then pop. It maintained aneasterly, level flight pattern for approximately four seconds, then began a series of 2-3clockwise, flat, wobbly, downward spins. It came out of the last spin, heading north, withthe nose at a 45 degree angle as it disappeared into the treetops."

Witness Imhoff, who was located approximately two miles northwest of the crash site ina residential back yard (looking northwest to south) reported seeing the plane during itsdownwind and turn to base leg. He stated that "the plane continued easterly, straight(and) level for a few (~20) seconds and then turned right, once again a generally flatturn, low speed, ~1500 feet elevation. As the aircraft came to a southeasterly heading, Iheard two loud pops about 4 or 5 seconds apart. Immediately after the second one, theaircraft began to roll to the right, going completely over twice, losing altitude, andgoing behind some tall trees, out of sight." Additionally, he reported that "asthe aircraft began to roll, with the left wing coming right, over the right wing, itseemed that the front to back axis of the aircraft was skewed to the right of thedirection of travel as if it were flying with the left wing ahead of the right wing. Itseemed to maintain this position through both revolutions and out of site."

Witness Biegel, an aircraft mechanic who was driving southbound on South Prairie Roadapproximately one mile north of the crash reported seeing both P-38's. He reported that"the silver lightning was in a slow turn in my direction" and that "Ithought its airspeed was a little slow and the aircraft went into a flat spin." Hereported that "ten seconds or less went by between level flight, the flat spin and(the) crash" and that "the aircraft was approximately 500 feet (in)altitude."

Witness Ryder, who was located approximately three miles north-northwest of the crashsite at the Tillamook airport (looking southeast) reported seeing a silver plane"falling out of the sky; not (in a) nose dive, just falling."

The pilot's father, who reported 10,000 hours of flight experience, of whichapproximately 3,000 hours of flight experience was in the P-38, also observed the aircraftduring the accident sequence. Witness Ervin Ethell, who was located approximately threemiles north-northwest of the crash site at the Tillamook airport (looking south) reportedseeing the aircraft "3 to 4 miles from the end of the runway" and "about 3to 4 hundred feet (in) height." He further reported that "It seem(ed) to me tobe much slower than what the normal approach speed should be" and that "I alsosaw the left wing drop pretty quick and the pilot immediately raise(d) the left wing up tolevel flight. Some 2 to 3 seconds later the aircraft made a slight right turn(approximately 5 or 6 degrees) and started down."

A private pilot flying into the Tillamook airport also witnessed the accident. WitnessFelley, who was flying a Cessna 172, announced via radio his intention to "perform across-field entry into the downwind leg for runway 31." He reported that "comingup to the airport's east side approaching hangar "B," I was descending throughapproximately 1,500 feet to get to pattern altitude and I noticed the P-38 a few miles tomy left, heading south." He further reported observing the aircraft transition from adescent to a climb "combined with what appeared to be a hard bank to the right, anearly 180 degree change in heading - perhaps the beginning of a roll (I believe it was tothe right), then it appeared to regain level attitude and proceeded immediately into aspin." He further reported that he believed that "the P-38 might have spun 4 to5 times in a slight nose-down attitude before it disappeared from sight" and that"it went in with little or no perceptible horizontal speed."

Witness Felley then diverted to the accident site, circling overhead at 1,500 feet for10 to 15 minutes and coordinated via cellular telephone with 911 providing instructions asto the location of the site before returning to land at Tillamook airport.

According to Tillamook County Sheriff's Dispatch records, the first telephonicnotification of the accident was received at 1800 hours and the first notification of adeputy on site was at 1825.

Personnel Information

Copies of six flight logbooks belonging to the pilot were reviewed. These logs spanneda time period which began in early spring of 1966, and concluded with a last entry ofNovember 11, 1996. While reviewing these logs it was noted that the pilot showed no loggedflights in the P-38. The flight times reported in the "flight time matrix" boxes(page 3, NTSB Form 6120.4) are based upon these six logbooks. These flight time tallies donot include any flight beyond November 11, 1996, up through the date of the accident. Theydo, however, include approximately 6-7 hours of reported P-38 flight time, all acquiredwithin the 90 day period prior to the accident.

Additionally, there was no recorded biannual flight review (BFR) noted within the sixthlogbook forward of June 1995. Without the pilot's seventh logbook, BFR action could not bedetermined. Additionally, the pilot held a second class Airman Medical Certificate datedJune 25, 1996, with the restriction that "holder shall wear correcting lenses whileexercising the privileges of his/her airman certificate." The pilot also held aStatement of Demonstrated Ability, dated September 29, 1994, for distant visual acuity inthe left eye and for lens replacement following cataract removal in the right eye."

According to the owner of N2114L (the olive green P-38 which was nearly identical tothe accident aircraft), the pilot flew his aircraft (N2114L) for a total of six to sevenhours during which (according to the owner) the pilot practiced simulated single enginemaneuvers. This was not, however, to include the actual in flight shut down and featheringof an engine, but only the reduction in power necessary to simulate an engine outcondition with a feathered propeller. During a portion of the flight time accrued on oneor more of these flights, the accident pilot (flying N2114L) was videotaped bothexternally and with an "in cockpit" video recorder. These sequences ofrecordings were compiled into a recording titled "Flying the P-38" lastingapproximately 30 minutes duration.

The pilot did not have a type rating in the P-38 series aircraft and his totalexperience in the P-38 aircraft was estimated up to but not including the accident flight(all in N2114L). The first time he flew N7973 was on the accident flight. The pilotpossessed a letter of authorization (LOA) provided by the Federal Aviation Administration(FAA) and dated July 21, 1987. The LOA opens by stating that "this letter authorizesyou to act as pilot-in-command in the following experimental category aircraft."

According to a letter to the wife of the owner of the accident aircraft, and signed byMr. Robert M. Barton, Manager, FAA Operations and Safety, Program Support Branch; anyauthorization to serve as pilot-in-command of the single seat P-38 aircraft registeredunder "limited" rather than "experimental" category would require atype rating in the aircraft.

Aircraft Information

The aircraft, a model F-5G (photo reconnaissance version of the P-38) serial number8006, was originally manufactured by Lockheed Aircraft Corporation in the early 1940's.According to records maintained by the FAA, the first indication of civilian registrationof the aircraft was documented via a bill of sale on March 7, 1946, and the aircraft(model F-5G) was registered as NX53753. The registration was subsequently changed toN503MH when the aircraft was sold on February 4, 1949. The next and final record of sale(to the current owner) was dated January 4, 1967. On June 6, 1970, the current ownerdeclared the aircraft un-airworthy and requested de-registration. The aircraft wassubsequently reregistered on November 27, 1996, as a P-38L by the current owner, and wasissued a new registration number of N7973. FAA records showed that the aircraft received aStandard Airworthiness Certificate dated "R" 05-07-58 as a P-38L for a"Limited" category.

According to the current owner, N7973 was being reconditioned to flying status byErickson Sky Crane of Tillamook, Oregon, and the intent was to place the aircraft in theSmithsonian Museum. One of the alterations to the aircraft as part of this process was theremoval of all of the bladder fuel tanks and their replacement with aluminum main andreserve tanks.

Three aircraft logbooks (airframe logs s/n 8006) were examined. The first log opened onMay 16, 1946, with a total of 28 hours 50 minutes brought forward from army records. Thelog closed with a final entry of April 19, 1950, with a total time of 834 hours 15minutes. The second log opened on April 20, 1950, and closed with a final entry of March12, 1954, with a total of 1,697 hours 10 minutes. The third and final logbook opened onMarch 13, 1954. The last recorded flight in the log was dated January 2, 1959, and closedout with 2,183 hours and 40 minutes of total flight time. The next entry in this log wasdated November 4, 1996, and documented the installation of the new left and right Allisonengines. The next logbook entry was an annual inspection which was dated January 8, 1997,and unsigned. The next logbook entry documented the installation of Winchester radiatorsand a commensurate revision to the aircraft's weight and center of gravity (CG), followedby a test run. This entry was dated May 30, 1997, but did not show the aircraft's totalempty weight or corrected CG. The next and final logbook entry was another annualinspection which was dated June 2, 1997, and signed off by Gary Austin, certified repairstation number JYDR439F. The total aircraft time at this last entry was recorded as2,188.7 hours. Total aircraft time at the time of the accident (4 days following theannual signoff) was unknown but estimated to be approximately 2,190 hours.

A single aircraft engine logbook for each Allison V1710 engine was examined. The rightengine, model V1710-111, was opened April 24, 1996, with a statement of major overhaul anda time since overhaul (TSO) of 6.0 hours. The last entry of the logbook (same page)reflected an annual inspection conducted on June 2, 1997, with a total time of 11.0 hours.This engine's propeller rotated clockwise when viewed from behind. The left engine, modelV1710-113, was opened October 13, 1995, with a statement of major overhaul and a timesince overhaul (TSO) of 6.0 hours. The last entry of the logbook (same page) reflected anannual inspection conducted on June 2, 1997, with a total time of 11.0 hours. Thisengine's propeller rotated counter-clockwise when viewed from behind.

Additionally, the Pilot's Manual (PM) states under "ENGINE FAILURE DURINGTAKEOFF" that:

And under "SINGLE ENGINE APPROACH AND LANDING" that:

The PM did not provide any information for aircraft minimum control speeds with theflaps fully retracted.

Aircraft Fuel System

The aircraft's original reserve, main, and outboard wing bullet-proof, bladder, fueltanks were removed during the reconditioning process. The left and right 55 (US) gallonoutboard wing tanks were not replaced and the fuel selectors had small vertical pinsinstalled to prevent the inadvertent selection of the unusable outboard wing tanks. Theleft and right reserve and main tanks were replaced with aluminum tanks. The capacity ofthe new metal reserve tanks was 44 (US) gallons each, and for the two new metal main tankswas 72 (US) gallons each. The aircraft's left and right metal drop tanks were left inplace, however, the plumbing to route fuel from these tanks to their respective engineswas not connected, thus the tanks were unusable.

The pilot/owner of N2114L, the olive green sister ship P-38 reconstructed earlier in1997, reported that the fuel consumption of his P-38 was nominally 60-62 US gallons/hourwith a minimum fuel consumption of 48 US gallons/hour best case and 120 US gallons/hourworst case for each engine. The specific engine flight chart for the P-38L equipped withAllison V-1710-111/113 engines showed a fuel consumption of 113 gallons/hour (2,600 RPMnormal rated maximum continuous power) and 63 gallons/hour (2,300 RPM maximum cruisepower).

Applying the 2,300 RPM cruise fuel consumption rate of 63 gallons/hour to the 44 gallonreserve tank capacity, yields a total fuel consumption time of approximately 42 minutesper engine from a full reserve (44 gal) tank. If the 2,600 RPM maximum continuous fuelconsumption rate of 113 gallons/hour is utilized, this yields a total fuel consumptiontime of approximately 23.5 minutes per engine from a full reserve (44 gal) tank.

According to the pilot/owner of N2114L, who was telephonically interviewed June 24,1997, both the left and right reserve and left and right main tanks of N7973 were toppedoff between 1000-1100 hours on June 6, 1997. He reported flying the accident aircraft forapproximately 20 minutes on the afternoon of June 6, and before the accident flight. Hereported that no fuel was added following this flight. He also reported that the normalprocedure for the operation of the aircraft was to takeoff with the individual engineselected to its respective reserve tank. Once established at altitude/cruise the left andright fuel selectors would then be set to main tanks. This procedure is described in the"Pilots Manual for Lockheed P-38 Lightning" which states under "NormalUse:"

and

It is not known whether the pilot/owner of N2114L, flew N7973 utilizing the reservetanks exclusively, or switched in flight to MAINS and then returned to RESERVE for theshort intermediate duration of the total flight (0.4 hours)

Meteorological Information

Aviation surface weather observations taken at Astoria (AST), Hillsboro (HIO), andNewport (ONP), Oregon, near the time of the accident reported the following conditions:

These observations were consistent with the description of weather conditions providedby the witnesses.

Wreckage And Impact Information

On site examination was conducted by an inspector from the FAA's Flight StandardsDistrict Office (Hillsboro, Oregon). The aircraft was determined to have crashedapproximately 50 feet east of the east side of a logging road in slightly up-slopingterrain. The latitude and longitude of the site was determined by plotting the site on achart, and was found to be 45 degrees 22.449 minutes north and 123 degrees 47.349 minuteswest.

The FAA inspector reported that the aircraft came to rest in an upright position withno lateral or longitudinal ground slide marks visible. The aircraft's longitudinal axiswas oriented along an east/west bearing (nose east). Additionally, he reported that asmall conifer was observed with minimal branch damage growing upright in an area boundedby the trailing edge of the main wing and the leading edge of the horizontal stabilizer(fore and aft) and the left and right tail booms. Additionally, a taller conifer with itslimbs sheared off and displaying substantial bark scraping was observed standing uprightalongside the right side of the right tail boom just aft of the wing. The rudder andelevator trim tabs were observed in a near neutral position. A small (2 inch diameter)tree was observed penetrating the right wing approximately six feet inboard of thewingtip, and a larger tree with considerable bark scraping along its trunk was noted atthe outboard trailing edge of the right. Both the right reserve and main tank gas capswere found secured at the site.

The inspector also reported that the aircraft's canopy was observed a short distancefrom the cockpit area and that the right propeller had separated from its engine. Oneblade was observed to be impaled vertically into the soil. The remaining two bladesdisplayed minimal deformation or scratching. The cockpit had been destroyed by fire,including the nose wheel tire and magnesium rim. The left propeller had separated from itsengine with the hub's longitudinal axis oriented approximately south. One blade wasseparated from the hub and impaled in the soil. The two remaining blades (one of whichdisplayed extensive twisting and bending deformation) remained attached to the hub. Theleft wing displayed a slight downward bend at a point approximately midway between theengine and tip and both the left reserve and main gas caps were found secured. The fuelselectors, located in the aft left portion of the cockpit were readable at the site andboth the left and right engines were observed to be set on "RES ON 44 GAL."

The FAA inspector also reported that all three landing gear appeared to be in aretracted position and that the Fowler flaps were retracted at the accident site.

Medical And Pathological Information

Post mortem examination of the pilot was conducted by Karen Gunson, M.D., at thefacilities of the Oregon State Medical Examiner, 301 NE Knott Street, Portland, Oregon,97212, on the morning of June 8, 1997. The report stated in part "GastrointestinalTract: The stomach contains approximately 200 cc's of partially digested, unidentifiablefood material-."

Toxicological evaluation of samples from the pilot was conducted by the FAA'sToxicology and Accident Research Laboratory and all results were negative.

Fire

The fire at the site was observed to be primarily focused within the cockpit area.Cockpit instrumentation, controls and switches were heavily damaged, therefore, themajority of Supplement "B" could not be completed. Combustibles within thecockpit area were hydraulic fluid in tanks located directly behind the pilot and themagnesium nose landing gear wheel rim, which, when the nose gear is retracted, is locateddirectly beneath the pilot.

Additional areas of less intense fire were observed within each engine's carburetorareas as well as the engine's turbo-supercharger units. Combustibles within these areasare characteristically, fuel, oil and hydraulic fluid.

The only other area of prominent fire damage to the aircraft was observed at the centerpoint trailing edge of the wing directly aft of the cockpit. A small semi-circular portionof the trailing edge of the wing and flap was burned away in this area. Combustibleswithin this area are characteristically, fuel (from the main tanks) and hydraulic fluidfrom behind the cockpit.

The soil beneath the aircraft, which sloped downhill from the wing trailing edgetowards the horizontal stabilizer and the road, consisted of dry porous dirt.

Survival Aspects

The Fire Chief who responded to the accident site reported that the pilot was observeda short distance south and east of the nose of the aircraft in a face down attitude andwith no evidence of thermal injury. The pilot's seat restraint system, including theshoulder harness and lap belt, as well as fragments of the seat pan and back, wereobserved in the same area. The pilot was still wearing his parachute and the aircraft'scanopy was observed on the ground near the pilot.

Tests And Research

The wreckage was re-examined on December 16/17, 1997. The entire horizontal andvertical stabilizer assembly was placed in a flat (fore and aft) attitude. When lookingforward from behind the surfaces both the left and right upper vertical stabilizers wereobserved to be bent towards the left. When the entire horizontal and vertical stabilizerassembly was placed in a 90 degree nose down (vertical axis) attitude and examined, boththe left and right lower (ventral) stabilizers were observed to be bent towards the right.

During the recovery procedure, the left wing was cut free from the airframe justoutboard of the fuselage. The inboard edges of both the reserve and main left wing fueltanks were examined. The reserve tank (forward of the spar and painted primer green) wasobserved to be generally deformed but displayed little fire damage. The main tank (aft ofthe spar and painted primer green) was observed to display some prominent damage at theinboard aft edge but displayed little fire damage. The tank also displayed a noticeableoutward deformation (hydraulic effect) and the seams along the top and bottom where foundto be split open.

The outboard magneto of each engine was observed to have sustained fire damage andthere was some crushing deformation to the under side of each engine. The inboard magnetoof each engine appeared to be in good condition. The carburetors and fuel filter screensfrom both engines were observed to have sustained some fire damage. Both the left andright engine fuel screens were removed an examined. Aside from a small amount ofcombustion byproducts, the filters were clear. The rocker box covers were removed fromeach engine and all rocker arms and push rods appeared to be in new condition. Airpressure was applied to a cylinder on each engine via a single spark plug socket. Duringthis application, the crankshaft of each engine was observed to rotate along withcommensurate rotation of the cam shaft and accessory gears. There was no observed evidenceof any mechanical malfunction with either engine.

The right propeller was examined and all three blades were attached to the hubassembly. One blade was smeared with dirt and the other two were relatively clean. Thesoiled blade displayed an approximate 20 degree sharp aftward bend from the plane ofrotation. The right propeller spinner was observed to display nearly equilateral upward(vertical axis) deformation on either side of the soiled blade, along with tensile typerivet separations toward the vertical axis.

The left propeller was examined and two of the three blades were attached to the hubassembly. One of the two retained blades displayed minimal deformation and was smearedwith dirt, while the other displayed "S" bending and blade twist. The detachedthird blade, which was also soiled, displayed some bending deformation. The left propellerspinner was observed to display deformation on opposing sides of the deformed and soiledblade, along with twisting diagonal pinch like deformation at some of its attach rivets.The spinner of this propeller was completely detached from the propeller hub.

The right Curtis electric propeller blade pitch drive motor was tested by applying 24volts of battery power to its drive motor leads. The unit remained attached to thepropeller mechanism and all three blades displayed a pitch change movement when power wasapplied. The blades were determined from the test to be in an intermediate pitch settingand not at the low (flat) or high (feather) stops. The left Curtis propeller blade pitchelectric drive motor was damaged to an extent that prevented testing.

Additional Information

Subsequent to the initial on-site examination, the wreckage was released to theinsurance representative for the purposes of recovery and storage. Recovery commencedapproximately 1400 hours on June 7, 1997, and was conducted by Mr. Harry Malette, H.L.M.Air Services, Inc. The wreckage was stored at the salvor's facility in Independence,Oregon. Written wreckage release was executed on December 20, 1997.

Probable Cause

Failure of the pilot to maintain minimum control speed (VMC), after loss of power inone engine, which resulted in a loss of aircraft control and collision with terrain.Related factors were: the pilot's improper fuel management and failure to change the fuelselector position before a fuel tank had emptied, which led to fuel starvation and loss ofpower in one engine; and the pilot's lack of familiarity with the aircraft, relative tosingle-engine minimum airspeeds.