WPR17FA128 · Prescot E. Wilkie RV-3

The airline transport pilot purchased the experimental, amateur-built airplane not fully assembled, and he subsequently completed the assembly. The Federal Aviation Administration (FAA) subsequently granted the airplane an airworthiness certificate in the experimental category. About 7 years later, the airplane was disassembled and shipped to the pilot's new home. The accident flight was the airplane's first flight after being reassembled.

A review of a video of the accident sequence showed that the left wing failed during a relatively benign maneuver as the airplane was descending. Several seconds earlier, the pilot had performed a high-speed, low-altitude pass over the runway with a sharp pullup. This maneuver would have produced much higher wing loads than those at the time of wing failure, and it is possible that the pull-up maneuver caused a localized buckling failure of the left-wing upper spar cap that further reduced its strength.

The accident airplane model has a history of in-flight wing separations. After a series of accidents, the kit manufacturer conducted extensive testing and provided owners/builders with suggested modifications for the airplane. The testing showed that the optional bonding of the spar caps together as a unit during assembly provided better resistance to buckling. The testing also showed that the initiating failure of the RV-3 wing was the buckling of the upper spar cap. The first modification to the RV-3 wing (Change Notice [CN] -1 issued in 1984) involved reinforcing the rear spar attachment point and strengthening the wing root rib. The second modification to the RV-3 wing (CN-2-I or CN-2-II, issued in 1996) involved the addition of stiffening angles to the main spar, which it deemed necessary for aerobatic strength.

Examination of the accident airplane revealed that the wing spar and carry-through had been modified by the addition of 0.125-inch-thick doublers on both the rear spar and center section. However, the rear spar attachment point was still a single-shear configuration. The wing root rib on both wings was 0.020 inch thick and had not been modified. This evidence indicates that the airplane had not been modified in accordance with CN-1. The FAA prohibited aerobatic operation of the RV-3 without the CN-1 modification.

The accident airplane was equipped with a Type I spar in accordance with its original design. No adhesive was found between the spar cap layers on the right or left wings. No modifications were made to the spar from its original design. Therefore, the accident airplane was not modified in accordance with CN-2-I.

The left and right wing attachment joint at the center of the airplane should have NAS 1304 or NAS 1305 bolts installed per the assembly instructions. Only three bolts of 10 bolts in the upper wing fittings and four bolts of 10 bolts in the lower wing fittings were NAS bolts. The type of bolt in two holes in the lower wing fittings could not be determined. The remaining bolts in the upper and lower wing fittings were AN5 bolts, which have a lower strength rating than the NAS bolts per the specifications. Most of the hardware installed in the upper and lower wing fittings was incorrect, and it was installed incorrectly, which significantly reduced the wing's tension capabilities. In addition, based on the condition of the left wing in the root area, it is likely that three bolts (those for upper hole 11 and upper and lower holes 10 and 11) were not installed in the left main spar caps when the airplane was reassembled.

Furthermore, examination of the left and right spar caps in the wing joint area revealed that the left upper spar cap was missing a 0.040-inch-thick shim, which would cause a mismatch in the spar cap stackup dimension at the airplane centerline where the wing fittings are installed. The mismatch would cause slightly abnormal loads in the wing fitting.

Lastly, no adhesive was found between the spar cap layers; therefore, the increased buckling strength that it may have provided was not achieved.

In summary, the omission of the bolts would produce a decrease in the buckling stability of the spar caps due to the lack of a clamping force holding the spar cap layers together. Adjacent to the missing bolts, the incorrect bolts were installed that had a significantly reduced tension capability further decreasing the buckling stability of the spar cap. Testing showed that the critical area of a correctly assembled RV-3 wing was the upper spar cap in the wing root area and the initiating failure was buckling of the upper spar cap. The wings on the accident airplane had further reduced strength due to the lack of adhesive, the missing bolts in the spar caps, and the incorrect bolts installed in the spar caps. The splaying and S-bending of the left wing upper spar cap layers around the fracture area is indicative of compressive buckling. The left wing failure likely initiated with a buckling failure of the upper spar cap in the wing root area.

Verbatim NTSB analysis and probable cause from the NTSB dataset