NTSB Factual Report

HISTORY OF FLIGHT

On January 21, 1995, at 1537 central standard time, a Questair- Venture, N4QV, operated by Condominium Insurance Specialists of America, Inc., of Arlington Heights, Illinois, impacted level terrain 3/4 mile southeast of runway 31R at Des Moines International Airport (DSM), Des Moines, Iowa, and was destroyed.

The two private pilot occupants received fatal injuries, and a post-crash fire occurred.

The flight was conducted under 14 CFR Part 91, and was en route from Palwaukee Airport (PWK), Palwaukee, Illinois, to Cheyenne, Wyoming, on an instrument flight plan. Visual meteorological conditions existed at the time of the accident. The airplane was cruising at 16,000 feet above mean sea level (msl), when the pilot declared an emergency and requested landing at DSM. During the emergency descent, the pilot stated he would be making a "dead stick landing."

Several witnesses described the airplane descending in a spin immediately before impact, and a fireball occurring after impact.

Post-crash inspection revealed no indications of engine rotation at impact. Disassembly of the engine revealed the crankshaft broken in two locations.

PERSONNEL INFORMATION

The 43 year old private pilot had approximately 2,450 hours of total flight time. He had privileges for single and multi engine airplanes, and an instrument airplane rating. He held a valid Third Class medical certificate, with no limitations. He was the designer, builder and primary machinist of the experimental powerplant. He held Aircraft Repairman certificates with the privileges of FAR 65.103 for engine assembly, machining and accessory, and for inspection of experimental aircraft make Questair Venture serial number 102, and Questair Spirit serial number 12.

The 60 year old private pilot passenger was the owner of the company which produced the Questair airframe kits. He had accumulated approximately 2,300 hours of flight time. His private pilot certificate included privileges for single engine airplanes and instrument airplane. He held a valid third class medical certificate with the limitations of "must wear corrective lenses while flying."

AIRCRAFT INFORMATION

The experimental airplane was a QuestAir Venture, serial number 43. It was equipped with a German made M & T, 3 bladed, composite propeller, model number MTV-9-D, serial number 93216. Each wet wing held 44.5 gallons fuel in 2 tanks, located in the inboard half of each wing, one in front of the other.

The airplane was equipped with PMA-550-TTV engine, serial number 1012. Based upon a Teledyne Continental Motors (TCM) 550- series engine, the PMA-550-TTV, is a twin turbocharged, 550 cubic inch displacement, 6 cylinder, fuel injected, reciprocating engine rated at 350 horsepower at 2,700 rpm. The accident engine was equipped with a combined electronic and single magneto ignition system. The engine was manufactured by Improved Aircraft Engines, Inc., dba Precision Made Airparts (PMA). The pilot was the chief designer and engineer for the engine.

The company dba, PMA, is not synonymous with the more commonly used term PMA as applied to manufacturers by the Federal Aviation Administration. The FAA issues a Parts Manufacturers Authorization (PMA) for certified aircraft parts. This does not apply to the experimental application of the Improved Aircraft Engines, Inc., components discussed in this report. Throughout this report, the term PMA will refer to the manufacturers dba, not an FAA authorization.

Designated as an experimental engine, it consisted of both modified components from certificated engines, and original PMA components (specifically the crankshaft).

At the time of the accident, only 3 technical persons were employed at PMA; the pilot, the pilot's son, and a third machinist. In building the PMA-550-TTV engine, Carillo connecting rods were utilized. The crankshaft connecting rod journals had to be machined to accept the smaller diameter rods, regardless of whether the crankshaft was from TCM or PMA. A PMA spokesman stated to the Safety Board that PMA had attempted to forge crankshafts, but had been unsuccessful. As a result, the process of completely machining a shaft from a solid steel billet was used.

According to PMA, no PMA manufactured crankshafts were produced until October of 1994. Further statements were made to the Safety Board that engine s/n 1012 was assembled and installed in N4QV in March-April 1994. Safety Board metallurgical and dimensional inspection of the failed crankshaft established that the shaft conformed to specifications provided by PMA, and did not match TCM crankshaft specifications.

The airplane owner agreement to purchase a PMA-550-TTV engine was drafted December 23, 1992, and accepted January 8, 1993. The construction work order for PMA-550-TTV engine s/n 1012 was opened on January 8, 1993. Invoice #952.868 dated August 9, 1993, showed the engine completed, with a partial payment due. A second invoice, dated December 24, 1993, showed engine s/n 1012 completed, with no payment due.

Records indicate that engine s/n 1012 was installed in aircraft s/n 43 on June 27, 1994. No record of the crankshaft having accrued operating time prior to this could be found.

All components installed appeared to be marked with the same serial number as the crankcase, ie; 1012. The exception to this was the crankshaft itself, which was identified as s/n 1008. Engine serial number 1008 was manufactured in the PMA facility during the period July 23, 1992, to December 1992, and delivered to its owner in February 1993. It was returned to the factory for inspection due to being inactive without preservation in January 1994. Engine s/n 1008 was disassembled, cleaned, and reassembled between January and March 1994.

The periods of overlap at the factory between engines 1008 and 1012 were December 1992, to February 1993, and again between January and March 1994. It could not be determined if both engines were actually disassembled at the same time. Disassembly and inspection of engine s/n 1008 in April 1995, revealed a modified TCM crankshaft with no identifying serial numbers visible.

The main landing gear and flaps were operated by a common electric motor, and the nose landing gear was operated by an independent electric motor. Electrical power sources were the 24V 60 amp alternator and battery.

Serial number 43 was registered as having been built by Roger Bacon. An interview with the registered builder indicated that serial number 43 had been destroyed in an accident in Florida prior to 1994. During the interview, Mr. Bacon stated that the data plate he had attached to the airplane he built was hand etched. The data plate, exhibiting serial number 43 which was attached to the accident airplane was machine stamped. Mr. Bacon further stated that he had an outboard piece of the wing from the airplane he had built, and that he did not know who had rebuilt serial number 43, or manufactured a new data plate. Registration records showed that serial number 43 was transferred from Roger Bacon to Condominium Insurance Specialists of America on March 11, 1992. It is unknown whether this was a transfer of the actual aircraft, or a paper transfer of the serial number.

Work order entries showed the following repairs having been completed by Windham Aviation, Inc., of N. Windham, CT, Repair Station #FUKR-043E, on June 27, 1994:

"Repair s/n 43 Questair Venture model M-20 by replacing the following assemblies: 1. Fuselage Fast Build Que-2000076-1 2. Right Wing Fast Build Ques-2010068-1 3. Left Wing Fast Build Que-2010068-2 4. Stabilizer Fast Build Ques-2020024-1 5. Elevator Fast Build Ques-2021023-1 6. Vertical Tail Fast Build Ques-2025021-1 7. Rudder Fast Build Ques-2026022-1 8. Canopy Tinted Ques-2006001-2

The same entry shows installation of a PMA-550-TTV engine, and the tach time being reset from 78 hours to 0.0 hours tach time. New flight test periods were requested by the owner on June 10 and July 6, 1994.

AIRPLANE HANDLING

Interviews with owners and pilots who have operated the Questair Venture indicated stall characteristics of approximately 58 knots indicated airspeed, a noticeable stall buffet prior to a moderate to gentle break, and falling off right wing down. No one was aware of the airplane having been tested for spin characteristics. The power off glide angle was described as very shallow with landing gear and flaps up, and much steeper (as much as 3 times steeper) with gear and flaps extended.

The airplane was described as very stable in either configuration. The only stall warning was aerodynamic buffet; no mechanical warning was installed. Good roll and rudder control throughout a stall was described.

WEIGHT AND BALANCE

Weight and Balance data derived from weighing on November 8, 1994 was as follows:

Basic Empty Airplane with full oil...1576.5 lbs at 67.72" Useable fuel...87.6 gallons at 86.0" Pilots at 101.0" Baggage at 124.0"

The Nominal CG Envelope provide in Questair Venture Manual Number 2091006, Vol 6, Revision 1.0, dated 7/14/93, shows a maximum weight of 2000 lbs. At this weight, the CG envelope is between 75.0 and 76.5 inches (Gear up), and 74.0 and 75.5 inches (gear down). This nominal envelope is not a limitation for an experimental airplane. The estimated weight at the time of the accident was 2445 pounds, at a CG of 77.24 inches aft of datum.

COMMUNICATIONS

The pilot declared an emergency with Minneapolis Center, and was handed off to Des Moines Approach Control while descending from 16,000 feet msl. DSM approach maintained radio contact and obtained landing clearance for the pilot from DSM ATCT. At 1523:07 CST, N4QV reported descending through 14,000 feet. The following transmissions were received from N4QV:

1523:21 "our battery is going dead...might lose radio contact." 1525:10 "had alternator alarm...the engine got rough but its smoothed out now...running on battery only so powers going down...we're going to extend gear shortly." 1525:50 "...leaving 11,600." 1527:37 "We have an alternator back on line. We're still going

to make a precautionary, but we've got power back." 1532:44 "Okay we have major problems now...we're coming down." 1532:52 "Okay it maybe a dead stick landing."

RADAR INFORMATION

Radar tracking data provided by Des Moines Approach Control shows the airplane in a constant spiraling descent from 14,300 feet. The spiral tends to decrease in diameter as the airplane descends. The rate of descent varies between 1,400 feet per minute and 1,000 feet per minute as the airplane descends from 14,000 to 5,900 feet MSL. During this time the radar reported groundspeed remains above 200 knots. As the airplane reaches approximately 5,900 feet at 1530:58 cst, the altitude momentarily stops descending, and the groundspeed begins to decrease steadily.

At 1532:44, at an altitude of 5,100 feet msl, N4QV transmits "we have a major problem now, we're coming down." From this point, the rate of descent increases to 1978 feet per minute, and the groundspeed continues to decrease to a minimum of 93 knots at 2,200 feet MSL at 1534:27 cst. This groundspeed occurs on the final downwind leg of the spiral approach.

At 1534:40 cst, the radar altitude track goes into "coast," and the last return received is at 1535:08.

WRECKAGE AND IMPACT INFORMATION

The airplane came to rest in the driveway of a small business, oriented on a heading of 075 degrees, 3/4 mile from the approach end of runway 31R. The left wing had a powerline wrapped around it, which appeared to have come from the power poles providing service to the adjacent building.

A single ground scar was evident which appeared to match the imprint of the left wing, and had a nose landing gear door at the base. This scar was located 17 feet from the main wreckage, and next to the power pole which supported the building supply wire. Debris was scattered in a clockwise direction from the wing scar to a point 120 feet in front of the main wreckage, and included the nose spinner, windshield glass, and red glass which appeared to match the position light on the left wing.

Both main landing gear were located on the right side of the fuselage centerline, and the nose wheel was located aft of the tail as the airplane was located.

The rudder, vertical stabilizer, and right flaperon separated from the airplane and came to rest to the left of the left wing.

The right wing was crushed upward along its inboard 1/2, exhibited a permanent 30 degree forward set, and was twisted trailing edge upward from 0 degrees at the root to 45 degrees at the tip. The wet wing ruptured and was completely burned.

The left wing exhibited a permanent forward set of approximately 5 degrees, was crushed uniformly aft 6 inches, and had no twist. The wet wing ruptured, with the inboard half burning, and the outboard half remaining unburnt.

Control rod and cable continuity was established. Both flaperons had separated from the outboard retaining pins. This appeared to occur when the wings were permanently deformed forward.

The 3 bladed composite propeller appeared to be at high pitch, with 2 blades intact, but splintered, and the third blade broken off at the hub, and located aft of the main wreckage.

Inspection of the engine revealed no rotational damage on either turbocharger. The magneto sparked. Fuel was present in the fuel lines to the firewall, and the fuel selector valve was open, with all lines clear and unobstructed. Oil was present in the engine, with a gold, shiny, metallic appearance. The crankcase itself was fractured in numerous locations forward of the number 3 cylinder. All accessories could be rotated by hand.

The gyroscopic instruments exhibited internal rotational scoring marks. Both intercoolers were clear and unobstructed. All 12 plugs were clean, dull gray and dry. All accessory drive gears were intact.

MEDICAL/PATHOLOGICAL INFORMATION

An autopsy was performed on both occupants by the Polk County Health Department/Deputy State Medical Examiner, Des Moines, Iowa, on January 23, 1995. Death was attributed to multiple blunt trauma. Toxicological results revealed no anomalies.

FIRE

A post crash fire occurred. The engine, fuselage, right wing, and left inboard wing exhibited extensive fire damage.

TESTS AND RESEARCH

The crankshaft, PMA s/n 1008 was inspected and tested for compliance with manufacturers specifications at the Safety Boards materials laboratory. In attendance was a representative from Teledyne Continental Motors. An invitation was extended to PMA, however, no representative was provided. Technical drawings were provided by PMA. Testing determined that the failed shaft conformed to dimensional specifications of PMA, and did not conform to dimensional specifications of TCM.

The shaft failed at two locations, both of which exhibited characteristics of fatigue. The primary failure location, between the number 3 connecting rod journal and the number 4 connecting rod journal exhibited fatigue characteristics which extended across the majority of the fracture surface. The exact origin was masked by post failure damage, however, appeared to be sub-surface (below the nitride layer).

The secondary failure location, between the number 3 main bearing journal and the number 4 connecting rod journal, exhibited mostly instantaneous overstress indications. This fracture surface exhibited an area where fatigue crack arrest lines were evident.

Visual examination of the number 4 crankpin bearing surface revealed no significant scoring marks or heat distress. Microscopic inclusions were noted in the uniform tempered martensite microstructure. All grain flow lines were oriented in the longitudinal direction of the crankshaft, indicating it had not been forged. The forging process would be expected to align the grain flow with the contours of the crankshaft.

Safety Board research could not determine the number of PMA crankshafts which had been produced. At least 29 PMA-550 series engines could be ident

NTSB Probable Cause

The pilot's failure to maintain aircraft control during the emergency descent, resulting in a stall/spin. Factors were material defects in the material used to manufacture the crankshaft, resulting in the crankshaft fracturing after approximately 200 hours in service.