NTSB Factual Report

HISTORY OF FLIGHT

On May 24, 1998, at 0630 hours Pacific daylight time, a factory-built kit, experimental airplane, Van's RV-8, N58RV, experienced an in-flight structural separation and crashed 1 mile south of Ripley, California. The aircraft was destroyed and the pilot and pilot-rated passenger sustained fatal injuries. The aircraft was being operated as a business flight by Van's Aircraft, Inc., under the provisions of 14 CFR Part 91 when the accident occurred. The flight originated from a private agricultural strip in Blythe, California, at 0618. Visual meteorological conditions prevailed at the time and no flight plan had been filed.

A relative of the passenger reported that, on departure, the pilot was seated in the front seat, while the passenger was in back.

The purpose of the flight was for the pilot to demonstrate the aircraft's flight characteristics to the passenger, who was believed by the kit manufacturer to be a potential kit buyer.

An agricultural pilot who was inbound to the strip in the opposite direction passed the southbound aircraft. He described the aircraft as being in straight and level flight at an estimated altitude of 500 feet agl. He said he did not notice anything unusual at the time.

An eyewitness, standing about 1.5 miles northwest of the crash site, reported that he heard the sound of an engine surging and looked in the direction of the sound. He saw a yellow aircraft flying straight and level, about 1,000 feet agl. The aircraft was on a southerly heading, about 1 mile east of his location. He estimated that he watched the flight for about a minute when he saw something fall from the aircraft. This was followed almost immediately by a loud boom that he described as sounding like a "shotgun." The aircraft's nose suddenly pitched up about 45 degrees then abruptly nosed over as it began a left roll. The aircraft entered a nose-down spin to the left, continuing in a vertical descent until impact.

A farmer, flying over the accident site a few days after the accident, reported that he observed crop discoloration in an arc-shaped pattern between the location of the left wing and the main wreckage.

PILOT INFORMATION

The pilot had been employed as a factory demonstration pilot for the manufacturer about 2.5 years. According to the kit manufacturer, a factory demonstration flight consists of turns, climbs, descents, slow flight, stalls and lazy 8's. There are no parachutes onboard and no aerobatic maneuvers are planned.

The pilot had built and flown his own RV-4 and had flown aerobatic maneuvers in that aircraft.

In addition, he was an experienced agricultural pilot and a commercially rated rotorcraft pilot, having served two tours in Vietnam as an Army helicopter pilot.

AIRCRAFT INFORMATION

The aircraft was the second factory-built kit aircraft of this design, having been completed on March 28, 1997, and received its airworthiness certificate on April 2, 1997. It had flown over 400 hours since that time. As part of its flight test program, the aircraft had performed loops, rolls, immelmann turns, horizontal eight's, spins, and other similar positive g aerobatic maneuvers without incident. The kit was designed to meet the design standards of 14 CFR Part 23.

The aircraft was designed for flight loads that were within +6 g's and -3 g's. The wing design has been static tested to +9 g's. Since the wing spar is nearly symmetrical in cross section, there were no negative g static tests or flight test maneuvers performed. The maximum allowable gross weight is 1,800 pounds; however, the maximum allowable gross weight for aerobatic performance is 1,550 pounds. A g-meter was not installed at the time of the accident nor were there parachutes on board.

The wing spar specifications are for 2024T3 sheet (web), 2024T351 (caps) and 2024T3 sheet (stiffeners).

The aircraft was equipped with dual controls; however, the brakes can only be operated from the front seat. The elevator trim is actuated by an electric servo, which the pilot controls with a 4-way switch that was mounted on top of the front seat control stick.

The aircraft was equipped with an electronic engine management system that has a nonvolatile memory.

A newly designed right-wing aileron was installed on May 8, 1998. The design change was intended to decrease burbling that had been experienced with full deflection control movements. A flight test that was completed on May 15, 1998 included full control deflection up to the maneuvering speed of 142 mph. No adverse effects were noted during the test. The aircraft had flown about 8 to 9 hours since the installation.

The aircraft is equipped with two main fuel tanks, each with a capacity of 21 gallons. There is no flow through between the tanks. The fuel selector has three positions: left, right, and off. There was one inverted pickup tube in the left tank. The tanks are vented separately.

The aircraft's fuel system is equipped with a fuel boost pump that is intended to remain on during takeoff and landing.

The aircraft was last refueled on May 23, 1998, at Payson Aviation in Payson, Arizona. At that time, it was serviced with 31.4 gallons of 100-octane low lead aviation fuel.

METEOROLOGICAL CONDITIONS

The localized weather conditions at the time of the accident were uniformly described by witnesses near the crash site as being calm and clear.

WRECKAGE AND IMPACT INFORMATION

A postaccident inspection of the aircraft by the Safety Board found an outboard section of the left wing about 0.17 miles northeast of the main wreckage on a bearing of 240 degrees to the main wreckage site. The coordinates were 33 degrees 30.018 minutes north longitude and 114 degrees 38.734 minutes west latitude.

The main spar of the left wing was fractured at a point inboard of the aileron and outboard of the flap. The main spar of the right wing was also fractured about the same location, but remained attached by the wing's outer skin.

Fragments of the canopy were found between the separated wing and the main wreckage. Each fragment was examined for evidence of foreign material transfers; however, none were found.

The main wreckage was located at 33 degrees 29.967 minutes north longitude and 114 degrees 38.934 minutes west latitude. The elevation of the accident site was estimated at 380 feet msl.

The aircraft was found in an agricultural field, crushed, and buried in the ground to a depth of 5 to 6 feet so that only the empennage and the remaining wing structure were visible. The leading edges of both horizontal stabilizers were crushed. The vertical stabilizer had "scorpioned" over, with crushing visible on its leading edge. The remaining left wing structure and the right wing showed leading edge damage along with aft bending of the spars.

The elevator trim tab control arm was separated at the clevis from the control surface. The trim actuator control rod was extended 1.5625 inches (full extension) tab up. The arm has a clevis on both ends. The clevis is made of DuPont nylon, a composite material, comprised mainly of nylon and fiberglass. The rod is threaded with a jam nut fixing the position of the clevis. The control arm itself is made from stainless steel. The control arm had separated at the aft clevis. Further examination revealed a notch had been cut, making the clevis deeper. After a discussion with the kit manufacturer, it was established that the extended notch was not a design feature. However, without the extended notch, it was evident that the clevis would have contacted the control horn on the elevator trim tab during it range of movement.

The aircraft was removed from the impact point with a backhoe, in order to recover the occupants. This prevented Safety Board investigators from examining the wreckage while it was still in an undisturbed condition. Photographic evidence provided by the Riverside County Sheriff's Office was the basis for the wreckage condition and impact description.

The wreckage was recovered to an inspection site in Phoenix, Arizona, and laid out in a two-dimensional manner on May 27, 1998. All control surfaces were identified. Due to the degree of impact damage, it was not possible to establish control continuity.

The forward control stick was located; however, the electric elevator trim switch was destroyed. Engine control continuity was established from the front seat. Rear seat controls, though destroyed, were identified as being installed.

The fitting connecting the right tank fuel line to the fuel selector was less than finger tight and was separated from the selector with less than three full turns. The fuel selector valve exhibited multiple internal fractures. The selector handle was displaced but remained closer to the left position than it was to the off or right position. No fuel selector position is specified in the aircraft POH for takeoff or landing but the kit manufacturer suggested that it is common practice to select the fuller of the two tanks during those phases of flight.

An engine teardown was also conducted on this date. When the No. 1 and 2 cylinders were removed from the case, it was noted that the front of the case was loaded with compacted dirt. The cooling fins to both cylinders exhibited aft crushing. Crankshaft rotation was limited to about 20 to 30 degrees with binding. Disassembly continued with the remaining cylinders being removed. When the crankshaft was removed, it was visually noted that there was an apparent bend at the No. 1 rod journal. The oil sump was fractured into several pieces. The ring gear and support housing had fractured into several pieces. The rear crankshaft gear was in place with the safety intact. The top of the engine case was cracked but there was no evidence of an oil leak. The engine data plate was not recovered.

The top spark plugs, rocker box covers, and accessory case were removed. The rocker box covers, while in place, all exhibited some degree of crushing and denting. With the limited crankshaft rotation available, some movement of the rocker arms was noted. All of the push rods and housings were in place. The No. 1 and 2 push rods were crushed, while the remaining housings showed varying degrees of damage. The oil return lines were separated except for the No. 1 line.

The No. 2, 3, and 4 top plugs were in place with the lead wires attached; however, the No. 1 plug had separated above the hex fitting. The bottom No. 1 plug was in place with the lead wire still attached. The No. 3 plug had separated at the hex fitting while the No. 4 had separated just above the hex fitting. The No. 2 plug had been pulled out of the cylinder and was not recovered. With the exception of impact damage and oil soaking, the wear and coloration on the plug electrodes was consistent with normal operation according to the Champion Spark Plugs Check-A-Plug chart.

The accessory section was broken open at the lower left corner. All of the rear accessories had separated and exhibited varying degrees of damage. The left magneto was capable of being rotated by hand. The impulse coupling functioned and it produced a spark from all four lead wires. With the accessory case removed, the oil pump could not be rotated by hand. When it was opened, dirt was found in the body. Both of the internal steel impellers were intact.

The carburetor was detached and separated into two sections. The floats and venturi were not located. The intake air box, while crushed, was still attached to a section of the carburetor with the filter element still in place. The exhaust and intake tubes were separated and crushed.

The propeller was still attached to the propeller flange. Both blades were still attached to the hub although one blade rotated in the hub. Both blades were bent back about 90 degrees along the cowling and exhibited a twist. The spinner was separated and one blade exhibited several leading edge gouges.

MEDICAL AND PATHOLOGICAL INFORMATION

An autopsy was conducted on May 27, 1998, by a county contract pathologist for the Riverside County Coroner's Office, with specimens retained for toxicological examination. The toxicological test results were negative for alcohol and all screened drug substances.

TESTS AND RESEARCH

The fractured wing spars from both wings were submitted to Seal Laboratories in El Segundo, California, for examination. The fracture surfaces were inspected visually and with an optical stereomicroscope. Some fracture surfaces were examined with a scanning electron microscope (SEM). Rockwell hardness was measured on the flat side of spar. Electrical conductivity was measured using an eddy current. A cross section of the spar was mounted in epoxy, mechanically polished, and etched to reveal cladding. A sample from the spar was analyzed for elemental chemical composition by emission spectroscopy. The spar surfaces were analyzed for elemental chemical composition in a scanning electron microscope with an energy dispersive X-ray microprobe attachment containing a "thin window" detector. The EDX microprobe spectra were obtained at 20 kV and 5 kV electron beam voltages, for detection of all elements above the atomic number 5 (boron).

The metallurgist reported that the left wing spar had evidence of a ductile fracture due to a positive overload. The right wing spar had evidence of a ductile fracture due to a transverse (fore to aft) overload. The spar material met design specifications for metal composition and hardness. There was no evidence of fatigue or corrosion found in either spar fracture. The outboard section of the left wing did not exhibit any lateral rippling that could be associated with pre-failure aeroelastic divergence.

The tensile and compressive resistance was measured on exemplar elevator trim actuator control arms, both with and without an extended notch being cut in one clevis end. The tensile failure occurred at 382 pounds and 402 pounds without the extended notch. It occurred at 413 pounds with the extended notch. In all three cases, the clevis separated.

Compression testing was performed at Durkee Testing Laboratories, Inc., in Paramount, California. The compression failure occurred at 160 pounds (shaft bent) and at 382 pounds and 402 pounds when the clevis separated.

The air load on the elevator trim control arm at 160 mph was calculated by the kit manufacturer as 27.07 pounds. At 190 mph, the load was calculated as 40.93 pounds. The stick forces that the pilot would need to exert in order to neutralize the down loads generated on the elevator in level flight were calculated as 24 pounds at 160 mph and 29 pounds at 190 mph.

The trim actuator was functionally tested and inspected at Menzimer Aircraft Components, in Vista, California. The trim relay passed a continuity check. The actuator ran full travel in both directions drawing .18 milliamps. The actuator was opened and the proper lash and alignment of the nylon gears was verified. The actuator is rated at 100 pounds. The manufacturer stated that it is not possible for the control rod to have been driven to the position in which it was found by crash forces. In order for it to move, he said it must be supplied with electrical power, a closed circuit, and a directional switch. The time to slew from neutral trim to full nose down was 6.5 seconds.

Aeroelastic divergence (flutter) analysis for the main wings was performed by Latoni, Inc., in Vero Beach, Florida. In preparation for the analysis, ground vibration test (GVT) was conducted in North Plains, Oregon. The test subjects included a wing with an empty fuel cell and repeated data collection with a full fuel cell. The data obtained from the GVT formed the basis of the flutter analysis.

Wing flutter analysis found no flutter within the aircraft's flight envelope. The analysis included a simulated failure of the aileron control rod.

Horizontal tail flutter analysis found no flutter within the aircraft's flight envelope. The analysis included balanced and unbalanced conditions, as well as fixed stick and free stick modes.

Elevator trim tab flutter analysis found no flutter within the aircraft's

NTSB Probable Cause

the intentional or unintentional sudden application of aft elevator control by an undetermined aircraft occupant that exceeded the design stress limits of the aircraft. The aircraft gross weight, which exceeded the maximum allowable for aerobatics, and airspeed, which exceeded the maximum maneuvering speed for the weight, were factors in this accident.