NYC00FA248
NYC00FA248

HISTORY OF FLIGHT

On September 3, 2000, at 1346 eastern daylight time, a Piper PA-32R-301, Saratoga SP, N8230G, was destroyed during an in-flight breakup over South Kingstown, Rhode Island. The certificated private pilot, the pilot-rated passenger, and another passenger and her dog were fatally injured. Instrument meteorological conditions prevailed, and an instrument flight rules flight plan was filed for the flight, between Chatham Municipal Airport (CQX), Chatham, Massachusetts, and Northeast Philadelphia Airport (PNE), Philadelphia, Pennsylvania. The personal flight was conducted under 14 CFR Part 91.

According to Federal Aviation Administration (FAA) records, the flight departed Chatham about 1305, and was level at 8,000 feet when the pilot contacted Providence Approach Control at 1329. During that contact, the controller asked the pilot if he would like a more southerly routing, and the pilot responded that he would. About a minute later, the controller gave the pilot new routing, and the pilot read back the clearance. At 1337, the controller told the pilot to change to another radio frequency, and at 1338, the pilot checked in with the next sector's controller. No further transmissions were received from the pilot; however, a sound similar to that of an emergency locator transmitter (ELT) was heard in the background when the controller spoke to another aircraft at 1346. During a review of the air traffic control audio tape, no unusual background sounds were heard during the pilot's transmissions, and the pilot never advised the controllers of any difficulties.

A witness on the ground, a former Navy pilot, stated that he was out in the yard, less than a mile away from the accident site at the time of the accident. There was an overcast sky, about 100-200 feet above the ground, and the winds were "pretty near calm."

Although the witness did not see the airplane, his attention was drawn to it because of a sound, which was similar to the pilot increasing the prop to full rpm. The sound came from about 45 degrees above the horizon, but there was no doppler shift, and the airplane didn't sound as if it was circling or spinning. The sound's bearing did not change, but its azimuth decreased.

The witness then heard a "pop," and another sound, like "something was coming off of the airplane." He heard the engine "rev up," followed shortly thereafter by silence for about 5 seconds, then the sound of an impact.

The witness also noted that when he heard the engine, it was running smoothly and was not sputtering.

According to transponder and radar information, the airplane was level at 8,000 feet, proceeding southwest, then made a turn to the west at 1342. About a minute later, it turned back toward the southwest. Around 1345, the airplane began a turn to the right, which continued through about 270 degrees of arc. It then turned back to the left for about 30 degrees, and remained headed eastbound for about 15 seconds before radar contact was lost.

Altitude information indicated that the airplane had maintained 8,000 feet until 1335:21, when it began a descent. It leveled off at 7,400 feet at 1337:21, then began a climb, and reached 8,500 feet at 1341:12. It maintained 8,500 feet for about a minute, then descended to 8,200 feet, and maintained between 8,200 feet and 8,100 feet until 1345:26. At 1345:30, the airplane was at 8,000 feet. At 1345:39, it was at 7,800 feet, and at 1345:48, at 7,000 feet. Additional altitude readouts were: 6,200 feet at 1345:53; 3,800 feet at 1346:02; and 1,900 feet 1346:12.

Correlating approximate altitude and heading information, the airplane began to descend shortly after beginning its 270-degree turn to the right. It passed through 7,000 feet just before reaching 180 degrees of turn, and was at 5,200 feet when it passed through 270 degrees. The first 180 degrees of turn were completed in about 30 seconds, while the last 90 degrees were completed in about 7 seconds. Approximate rates of descent were calculated as follows: between 8,000 and 7,800 feet: 1,300 feet per minute (fpm); between 7,800 and 7,000 feet: 5,300 fpm; between 7,000 and 6,200 feet: 9,600 fpm; between 6,200 and 3,800 feet: 16,000 fpm; between 3,800 and 1,900 feet: 11,400 fpm.

The accident occurred during daylight hours, and the main wreckage was located at 41 degrees, 26.25 minutes north latitude, 71 degrees, 32.51 minutes west longitude.

PERSONNEL INFORMATION

The pilot held a private pilot certificate, was rated in single engine land airplanes, and held an instrument-airplane rating. His latest FAA second class medical certificate was dated December 9, 1997.

The pilot's partially recovered logbook listed 1,291 total hours of flight time as of August 13, 2000. The pilot also recorded 195 hours of actual instrument time and 127 hours of simulated instrument time. All flights listed in that logbook, which began in late 1997, were flown in the accident airplane. The pilot completed a flight review and an instrument proficiency check on April 17, 2000.

According to training records, the pilot received training at Recurrent Training Center, Inc., Savoy, Illinois, in September 1998. At the time, he reported 1,170 hours of flight time, with 168 total instrument hours and 824 hours in the accident airplane. Training sheets from that time indicated that the pilot completed four simulator flights, but there was nothing noted regarding partial panel training.

According to additional training records, the pilot also attended training at Aviation Training Management, Inc., Vero Beach, Florida, in April 2000. He received 6.7 hours of training, including partial panel (IFR) training.

The pilot-rated passenger held a private pilot certificate, was rated in single engine land airplanes, and held an instrument-airplane rating. She reportedly had not been actively flying for a number of years. Her latest FAA third class medical certificate was dated July 31, 1986.

METEOROLOGICAL INFORMATION

About the time of the accident, an airport approximately 10 nautical miles to the northeast reported calm winds, visibility 2 statute miles, mist, temperature 74 degrees Fahrenheit, dewpoint missing, ceiling 500 feet above the ground, and an altimeter setting of 29.95 inches.

A pilot who departed Providence at 1410, reported that he was "skimming" cloud tops while proceeding southward at 5,000 feet. He also reported that he entered a cloud buildup around 1415, about 12-15 miles south of Providence, and remained in it for about 5 minutes. Conditions inside the buildup included heavy rain and moderate turbulence, which was "uncomfortable but not severe."

A radar image taken 16 minutes before the accident revealed light to moderate precipitation along the southern Rhode Island coastline, but nothing in the vicinity of the accident.

AIRPLANE INFORMATION

The Piper Saratoga was constructed in 1980, and powered by a Lycoming IO-540 engine. According to the airplane's registration certificate, the pilot had owned it since February 1984.

Maintenance logbooks were not located. According to work orders, the last annual inspection was completed on December 7, 1999. At the time, the hour meter indicated 1,458.6 hours.

The airplane was equipped with a Parker Hannifin Airborne model 211CC dry air pump and a Precise Flight, Inc. standby vacuum system.

WRECKAGE AND IMPACT INFORMATION

The main wreckage consisted of the fuselage forward of the aft-most seats, including the engine and propeller, parts of the right wing, and both passenger doors. The wreckage was located in a small clearing only slightly larger than the airplane, and there was almost no damage to any of the surrounding trees. The main fuselage was pointed toward 130 degrees magnetic, and the engine compartment was still attached, but offset to the left, about 20 degrees.

About 1,900 feet directly north of the main wreckage, was the right half of the stabilator. About 200 feet to the south of it was the left half of the stabilator. About 100 feet further south of that was the rudder, and near it, the vertical fin. To the east of the vertical fin, about 250 feet, was the aft left portion of the fuselage. To the southwest of the vertical fin, about 500 feet, was part of the inboard section of the right wing. About 400 feet southwest of that was the left wing, and another 200 feet further south, was the cargo hatch.

The wreckage and debris were removed, and laid out in a field where all flight control surfaces were accounted for. The cockpit had been destroyed; however, cable continuity to the tail and right wing was confirmed. Control cables to the left wing were separated, with a "broom-straw" appearance. The landing gear was down, and flap position could not be confirmed due to impact damage.

The stabilator spar was separated about mid-span, with the lower cap remnants protruding downwards on both sides of the break. The vertical fin was dented approximately halfway up the leading edge, but was not twisted. The left wing spar exhibited downward bending of the lower spar cap where it had separated.

One of the three propeller blades was buried in the ground, up to the propeller hub, approximately vertical to the ground. When the engine was lifted, the blade was found to have been crushed and folded against a rock. Other than the outboard 4-5 inches, there was no leading edge damage. The other two propeller blades were undamaged. All propeller blades were in a "high pitch" position, and the propeller governor was set for high pitch, as well.

The engine was rotated after fractured parts of the lower casing were removed. Drive train continuity was confirmed, and compression was verified in all cylinders. No fuel was found in any of the lines, but a receipt indicated that the airplane was refueled with 35 gallons of fuel at Chatham Airport. The alternator face plate did not exhibit any rotational scoring, but did exhibit vertical scoring. The magnetos produced spark on all leads, and the upper spark plugs were light gray in color. The vacuum pump shaft was sheared. The air-driven directional and attitude gyros did not exhibit rotational scoring, but the electric-driven turn coordinator gyro did. The airplane was equipped with a standby vacuum system, connected to an engine air induction tube. Proper alignment of the vacuum source fitting to the induction tube could not be verified due to impact damage.

MEDICAL AND PATHOLOGICAL INFORMATION

On September 5, 2000, an autopsy was performed on the pilot's remains at the State of Rhode Island, Department of Health, Office of Medical Examiners, Providence, Rhode Island.

Toxicological testing was subsequently performed by the FAA Toxicology and Accident Research Laboratory, Oklahoma City, Oklahoma. Diphenhydramine was detected in the pilot's liver (0.14 ug/ml, ug/g) and kidney (0.042 ug/ml, ug/g).

According to several internet sites, diphenhydramine was an antihistamine found in products such as Benadryl. Common side effects included dizziness and drowsiness.

TESTS AND RESEARCH

-- Vacuum Pump Service Letter --

According to an Airborne Service letter, dated August 16, 1991:

"Environmental conditions such as heat, solvents, oils, etc. can shorten the useful life of the urethane coupling. In order that air pump operation not be interrupted, we recommend the coupling assembly (including spline drive) be replaced every six (6) years with a new date-coded Airborne coupling assembly. This coupling can be easily replaced by any authorized mechanic. This feature is unique with the Airborne dry air pump. Coupling assemblies utilized in pumps and all spares manufactured after August 1, 1976, are date (year) stamped on the coupling face."

-- Vacuum Pump and Coupling --

According to a diagram provided by Parker Hannifin, the part of the vacuum pump system that fit into the engine was the "coupling drive end." The coupling drive end connected to the "drive end" of the flex coupling, which then connected to the "driven end" of the flex coupling. The driven end of the flex coupling connected to the "coupling driven end," which, in turn, connected to a shaft assembly.

The vacuum pump and couplings were examined at the Safety Board Materials Laboratory. According to the factual report, examination of the pump revealed that the driven end of the flex coupling was still attached to the driven end coupling. The driven end coupling was found to rotate freely.

The torque indicators on the nuts and bolts attaching the flanges to the housing were intact. On the mounting flange adjacent to the "IN" tube was an identification tag. Approximately half of the identification tag was missing but the remaining section indicated that the pump was an "AIRBORNE DRY AIR PUMP", "MADE IN USA" by "AIRBORNE MFG CO." of "ELYRIA". Airborne representatives identified the pump as a model 211CC.

Inside the pump housing, there were six vanes. Each of the vanes was free to move within its slot. There was a circumferential rubbing band on the periphery of the rotor.

The rotor was removed from the housing. Light circumferential score marks were observed on the outside diameter adjacent to the rubbing band. Rubbing marks were also observed on the edges of all six vanes. The vanes were removed and the thickness, length and width were measured.

The inside surface of a quadrant of the housing had rubbing marks parallel to the rotor center, commonly referred to as a "washboard effect." Surface coating had been removed, revealing the material underneath. The rubbing marks, accompanied by light circumferential scoring, were more prolific in one area; however, similar features were observed in the other quadrants.

A portion of the flex coupling was removed from the driven coupling in one piece. Visual examination of the fracture area revealed that the fracture had multiple planes, primarily in a conical shape. A portion of the fracture surface had been smeared circumferentially.

On the backside of the driven end of the flex coupling, the letters "AW", the date "1979", and the number "1" were cast into the material.

The fracture face of the drive end of the flex coupling was "also largely on a conical plane." The drive end of the flex coupling was removed in two pieces. The fracture between the two pieces consisted primarily of two spirals that intersected the coupling center. The two spirals also intersected two diametrically opposite drive pin holes. Cracks similar in location and direction to those [above] were also observed on four of the remaining six drive pin holes.

Fracture faces of the spiral fractures, originating next to the pin holes, were examined. Fracture lines emanated from inside diameter of the pin hole along the fracture.

The fracture surfaces, along with fractures that emanated from the other holes, intersected the holes at a 90-degree angle, "approximately at the 7:30 position, with 6:00 being the inboard side of the hole, and 12:00 being the outboard side of the hole. The portion of these fractures adjacent to the holes was very flat, typical of a brittle fracture mode. The spiral direction of these cracks and the location of the initiation area on the drive pin holes were typical of brittle cracking as a result of torsional loading of the flex coupling."

Hardness testing was conducted in accordance with ASTM D2240 Standard Method, using the Shore D Scale. It was performed on one of the drive end flex coupling pieces and on two other couplings supplied by Airborne. The drive end flex coupling piece displayed readings of 49, 50, 49, 50 and 49 (an average of 49.4). A 1988 coupling displayed readings of 50, 50, 51, 51 and 50 (an average of 50.4), and a 1998 coupling displayed readings of 50, 49, 51, 51 and 51 (an average of 50.4).

Examination of service letters revealed that the pump, model 211CC, should have been replaced every 1,100 hours and that the coupling assembly, consisting of the flex coupling and the drive coupling, be replaced every 6 years.

-- Standby Vacuum Supplement --

According to operating instructions excerpted from the flight manual supplement, under limitations:

"1. The standby vacuum system is for emergency or standby use only and not for dispatch purposes. 2. Vacuum powered and/or vacuum gyro-directed auto pilot operation may be unreliable when the SVS is the sole source of vacuum. Vacuum powered or vacuum gyro directed autopilot should be OFF when operating with failed primary vacuum system. [3. Refers to operation with de-ice systems.] 4. Above 10,000 feet pressure altitude, engine power settings may have to be significantly reduced to provide adequate vacuum power for proper gyro instrument operation."

Under emergency procedures:

"1. In the event of (warning light) primary vacuum system failure, turn the standby vacuum valve handle to LEFT-RIGHT alignment (ON) and reduce throttle setting as required to maintain adequate vacuum power (suction gauge reading in green arc). If necessary, descend to a lower altitude to obtain a larger differential between atmospheric pressure and engine manifold pressure. Vacuum power must be closely monitored by checking vacuum gauge frequently. 2. CONTINUED IFR FLIGHT IS NOT RECOMMENDED AND IMMEDIATE ACTION SHOULD BE TAKEN TOWARD VFR CONDITION OR LANDING. 3. If descent is impractical: a. Periodically reduce power to 'spool up' the gyros. b. Reapply power as required while comparing vacuum driven gyros against the turn and bank, turn coordinator, VSI, and other flight instruments, and c. When an obvious discrepancy is noted between the vacuum driven instrument and the other flight instrument REPEAT the above 'spool up' procedure as needed."

-- Additional Vacuum Items --

In October 2001, an examination of the airplane by third parties occurred at the facility where the wreckage was stored. Several items were wrapped, and subsequently forwarded to the Board for further examination. Notable items included:

Low vacuum annunciator light (Piper): The light bulb filament was examined under magnification, and found to be stretched.

Low vacuum annunciator light (Precise) the "INSTRUMENT SOURCE PUMP INOP WARNING": No filament stretch could be examined because the light was a diode-type. The light was, however, connected to a 12-volt battery and found to be operational.

Auxiliary vacuum off-on valve (Precise): The handle had been broken off, and the ball valve was partially closed, about 1/8 of an inch. One of the two outside surfaces of the ball valve had a scratch, approximately perpendicular to the lip of the valve, also about 1/8 of an inch long.

-- Vacuum System Failure Simulations --

In response to a petition for rulemaking filed with the FAA to allow the substitution of a back-up attitude indicator for the turn coordinator, testing of human performance during vacuum system failures was performed by personnel from the Civil Aerospace Medical Institute (CAMI), Oklahoma City, Oklahoma. Testing results, as published in "Proceedings of the Human Factors and Ergonomics Society 45th Annual Meeting - 2001," are located in the public docket.

-- Historical Data --

A review of available Safety Board accident data from 1983 to 2000 revealed that vacuum pump/system failures either caused, or were factors in, 30 accidents, resulting in 65 fatalities.

ADDITIONAL INFORMATION

-- FAA Pamphlet FAA-P-8740-52 --

According to FAA-P-8740-52, "The Silent Emergency,"

"All too often a pneumatic system failure leads to a situation where a pilot is forced into 'partial panel' instrument flying that he or she may not be fully prepared to handle."

In addition,

"A complete pneumatic system loss is noticeable immediately on the gage or within minutes by incorrect gyro readings. A slow deterioration may lead to sluggish or incorrect readings which may trap a pilot who is not constantly cross-checking all instruments - including the vacuum or pressure gage.

An additional factor involves an initial lack of recognition of the cause of the conflicting instrument indication which develops when one instrument indication which develops when one instrument, usually the attitude indicator, malfunctions. During training an instructor or safety pilot forces a partial panel scan by covering the attitude indicator. Although possibly proficient in flying 'partial panel,' many pilots are not trained or skilled in recognizing when to revert to a 'partial panel' scan.

If you are instrument rated and gyro instruments fail or mislead, do not be afraid to ask for help. ATC personnel know where to find better weather and are able to give 'no gyro' heading directions. Do not try to be a hero and continue on bravely as if a loss of pneumatic power was no big deal."

-- Minimum Equipment --

According to FAR Part 91, paragraph 91.205 (b), required flight equipment for visual flight rules flight included: airspeed indicator, altimeter, and magnetic direction indicator.

Paragraph (c) added a "source of electrical energy for all installed electrical and radio equipment" for night VFR flight.

According to paragraph (d), the following was included as required for IFR flight:

- Instruments and equipment specified in paragraph (b), and for night flight, instruments and equipment specified in paragraph (c).

- Gyroscopic rate-of-turn indicator, except on airplanes with a third instrument system usable through flight attitudes of 360 degrees of pitch and role.

- Slip-skid indicator.

- Sensitive altimeter adjustable for barometric pressure.

- Gyroscopic pitch and bank indicator (artificial horizon).

- Gyroscopic direction indicator (directional gyro or equivalent).

There was no requirement for a backup source of power for the flight instruments.

-- Mandatory Maintenance Requirements --

In a letter to the Director of Regulatory and Certification Policy, Aircraft Owners and Pilots Association (AOPA) dated April 25, 2001, the manager of the FAA's Small Airplane Directorate stated:

"Small airplane design approval holders cannot unilaterally impose mandatory compliance with manufacturer's SB's [service bulletins]. A statement that bulletin compliance is mandatory must be FAA-approved to be included in the Airworthiness Limitations Section of the Instructions for Continued Airworthiness (ICA). FAA policy does not permit this approval to be delegated to organizations or individuals."

In addition,

"FAA policy does not permit a predetermination that compliance with some future document is mandatory. Thus, SB's issued after the Airworthiness Limitations section is FAA-approved cannot be made mandatory without FAA involvement."

In an initiative written by the general manager of Parker Hannifin, dated February 22, 2001, Parker Hannifin requested that the FAA issue an AD which, among other things, would have limited single-engine piston aircraft to flight under daytime VFR, unless the aircraft had a "satisfactory and operational back-up pneumatic power source for the air-driven gyros, or a back-up electric attitude gyro instrument, to the pneumatically-powered gyro flight instrument system."

The letter also requested the "inspection and replacement of pneumatic system components in accordance with the aircraft manufacturers' and component manufacturers' instructions," as well as a requirement for an "independent and separate preflight operational check" of both the primary and back-up pneumatic source/back-up instrument.

In a response dated April 23, 2001, the manager of the Small Airplane Directorate stated that, regarding the limitation of single-engine piston aircraft to day VFR "unless equipped with dual pumps or backup electrical instruments," that an AD would be required for each type certificate. In addition,

"This AD could result in the FAA increasing the design standard that was established for these aircraft type designs. In addition, an AD that changes the type design operational limitations, will not correct the unsafe condition. If such a limitation were deemed appropriate, then a change to the Title 14 Code of Federal Regulations (14 CFR) Part 91 operating rules may be a more effective and efficient course of action. Potentially, Part 91 equipment requirements for IFR could be revised."

Regarding Parker Hannifin's requests for an AD for mandatory compliance with aircraft manufacturers' and component manufacturers' instructions, the Small Directorate manager stated that "the preamble to 14 CFR Part 39 specifically states that AD's are not to be issued to assure the use of normal maintenance practices or where individual cases of improper maintenance or lack of maintenance have caused a problem, unless the instructions are the source of the unsafe condition."

Regarding Parker Hannifin's request for an AD for certain preflight operational checks, the manager of the Small Airplane Directorate stated that it too, did not meet the criteria of 14 CFR 39, and therefore, the FAA could not issue such an AD. "The pilot is still responsible for determining that the aircraft is in condition for safe flight per 14 CFR Part 91.7."

The manager also stated,

"The FAA is evaluating methods to effectively reduce the occurrence of accidents caused by pilot spatial disorientation. We recognize the results of studies provided by industry. The FAA is examining loss of control in IMC and feels the factors leading to these accidents are those associated with the entire reference system and the system back-up. The system back-up is partial panel in many aircraft. These factors include: 1) failure of the vacuum pump, 2) failure of the attitude and directional gyro indicators, 3) ineffective annunciation of loss of vacuum system pressure and/or the gyro indicator failure, and 4) pilot confusion regarding the instrument indications whether failed or fully functional. Methods to reduce loss of control in IMC are currently being evaluated under the Safer Skies Initiatives. Evaluations from Safer Skies as well as other FAA initiatives could lead to recommendations to the Aviation Rulemaking Advisory Committee (ARAC) via the General Aviation Certification and Operations Issues Group for rulemaking. Recommendations for rulemaking must include a cost-benefit analysis as part of the justification. We would gladly accept industry support in these efforts."

-- Wreckage/Parts Release --

On September 7, 2000, the wreckage was released to a representative of the pilot's insurance company, United States Aviation Underwriters, Incorporated, New York, New York. The vacuum pump was forwarded to the representative on April 5, 2001. Additional parts, which were forwarded to the Safety Board in October 2001, were returned on November 29, 2001.

Use your browsers 'back' function to return to synopsis
Return to Query Page