CHI96LA041

The pilot reported that approximately 10 minutes into the flight a 'thump' was heard from the right rear of the airplane. He stated that besides a slight list to the right, he did not notice anything unusual with the airplane's flying characteristics. He decided to make a precautionary landing after noticing a rapid depletion of indicated fuel quantity on the upper wing tank gauge. After landing, the pilot and passenger exited the airplane and realized the airplane's upper right wing had departed the airplane. A fire started and consumed the entire airplane. The wing was found 7/8 mile from the resting point of the airplane. Examination of the failed upper right wing, revealed the wing had failed due to compression failure. The investigation also revealed that the failure was due to the presence of unbonded areas between the spar cap and the bidirectional fiberglass that secures the web to the cap. It is not possible to inspect these areas because they are not accessible after construction.

On November 24, 1995, at 1415 central standard time (cst), an experimental Sea Hawk, N391SP, was destroyed by fire after making a precautionary landing on a field in Marengo, Illinois. The pilot and passenger reported no injuries. The local 14 CFR Part 91 pleasure flight originated from Belvidere LTD Airport, Belvidere, Illinois, at 1355 cst without a flight plan. Visual meteorological conditions existed at the time of the accident.

In a detailed written statement, the pilot reported the airplane departed Belvidere and climbed to an altitude of 2,100 feet MSL. Once cruise flight was established, the passenger took the controls to conduct a series of turns to become familiar with the airplane. The pilot stated that approximately ten minutes into the flight, after the recovery from a 35 degree bank, a "thump" was heard from the right rear of the airplane. The pilot estimated that the maximum "g" load on the airplane during the recovery from the bank was less then two "g's."

The pilot reported he took control of the airplane and tried to determine the cause of the "thump." At the time, he believed that maybe a bird or duck had struck the pusher propeller. The pilot stated he exercised the controls along the roll and yaw axis and reduced the power. "Other than a slight list to the right which was compensated with slight left stick and rudder, there was nothing unusual in the feedback sensations through the stick, throttle or airframe," the pilot stated. He did notice a rapid depletion of indicated fuel quantity from the upper wing tank gauge. The pilot decided to make a precautionary landing on a harvested bean field.

The pilot reported the landing was relatively smooth with the airplane stopping approximately 500 feet after the initial touchdown point. The pilot and passenger egressed and began to assess the airplane's condition. It was at this time that the pilot realized the airplane's upper right wing had departed the airplane. A fire started and consumed the entire airplane. The upper right wing was found approximately 7/8 mile from the resting point of the airplane.

The airplane's log book shows the last inspection was an annual inspection performed on September 23, 1995. The airplane has logged 0.8 hour since the inspection. The airplane has logged a total 178 hours since its completion.

The inboard section of the failed upper right wing was sent to the NTSB Materials Laboratory, Washington, D.C., for examination.

The laboratory reported features of the composite materials of the wing in the aft portion of the lower spar cap and in the upper cap were typical of a compression separation.

Further examination revealed that the upper spar cap was partially separated from the bidirectional fiberglass tape which is used to secure the web to the upper cap. This partial separation extended over approximately 70 percent of the bonding area between the fiberglass tape and the upper cap. The report further states, "Almost all of the epoxy adhesive used to bond the lower surface of the upper spar cap to the bidirectional fiberglass tape remained on the tape. Examination of this adhesive showed that a large portion of its surface was rough, and recessed below other areas, and had a reflective appearance, indicating that the adhesive in this portion of the surface had cured without ever being in contact with the bottom surface of the upper spar cap." The bonded area between the upper spar cap and the spar web was also examined and revealed a continuous gap or unbonded area. This unbonded area was between the adhesive above the edge of the fiberglass layers and the spar cap in an area which contains no inspection ports so that the quality of the bond may be visually inspected. It was found on both the forward and aft sides of the spar. See attached Materials Laboratory Report.