FAIRCHILD AFB, WA, USA
N25AK
Mudry CAP 10B
During recovery from a left hammerhead stall, the pilot of a French-built Mudry CAP 10B transmitted that he had a locked control. At about that time, the airplane entered a steep right bank and began descending. It remained in a descending turn for three steeply banked 360 degree right turns, then it impacted the ground. Videotapes revealed rudder and elevator movements during the descending turn, but no aileron movement was seen from the time the pilot said he had a locked control until the airplane hit the ground. Investigation revealed two three-wire electrical bundles in each wing, which had been routed to the wingtip strobe lights and the wingtip smoke canister activation system. These created a resistance to movement of the aileron pushpull tubes. The right wing wire bundles had been routed through the center of the right aileron bellcrank mechanism. Testing revealed that the wire bundles in the right wing could become lodged in the bellcrank mechanism, and could jam the ailerons in a position that would cause the airplane to roll to the right. Flight tests simulating this condition in another CAP 10B confirmed that the airplane was not controllable when the ailerons were jammed in such a position. The manufacturer reported that the 1975 model airplane was not equipped with strobe lights or a smoke canister system when it left the factory. The investigation did not determine who had routed the bundles through the right bellcrank or for how long they had been in that location.
HISTORY OF FLIGHT On September 14, 1996, approximately 1300 Pacific daylight time, a Mudry CAP10B, N25AK, impacted the terrain during an aerobatic airshow performance at Fairchild Air Force Base, Spokane, Washington. The airline transport rated pilot, who was the sole occupant, received fatal injuries, and the aircraft was destroyed. The aircraft had been in the air for about ten minutes at the time of the crash. No flight plan had been filed, and there was no report of an ELT activation. According to Air Force officials, the pilot was participating in a two-day open house and airshow at Fairchild Air Force Base. He had been scheduled to perform his aerobatic demonstration once on Saturday, and once again on Sunday. The accident took place during the Saturday demonstration. According to witness statements, video camera tapes, and recordings of the tower communications frequency, the first indication that the pilot was experiencing problems occurred as he started a pull-out from a vertical descent after executing a left hammerhead stall/turn. During the descent, the aircraft began a slow clockwise rotation around the longitudinal axis, and by the time the pull-out was complete, the aircraft had rolled into a steep right bank of approximately 70 to 80 degrees. Near the completion of the pull out, while about 400 feet above the ground, the pilot made a radio transmission over tower frequency stating "I got a locked control." About seven seconds after that transmission, an unidentified person responded with "What did you say," and about one second after that, the pilot again stated "I got a locked control." At the time of this second transmission, the aircraft was established in the first of three steeply banked 360 degree right turns, during which the aircraft's bank angle almost constantly remained at more than 70 degrees. About five seconds after the pilot had transmitted the second locked control call, the aircraft reached a bank angle of nearly 90 degrees, and the pilot transmitted "I'm going to go in!" The pilot made no other transmissions after that, and the aircraft continued in a slowly descending steeply banked right turn, impacting the terrain about 35 seconds after the last transmission. As the aircraft descended, witnesses heard engine power changes, and when the aircraft was approximately 100 feet above the ground, all sound of engine power terminated. During the last 360 degree turn, the aircraft's airspeed appeared to decrease, and when it was about 50 feet above the ground, its nose appeared to pitch up slightly, followed by a rapid roll to the left. The aircraft impacted the terrain in a wings-level, 20 to 30 degrees nose-down attitude, while rolling to the left. During the investigation, videotapes of the flight were reviewed in both the stop-action and slow-motion modes, and it was noted that there was no observable change in the position of the ailerons from the time the pilot started the pull-out from the hammerhead until the aircraft impacted the ground. The video clearly showed that during this same period of time, both the rudder and elevator were almost constantly being moved to different positions. The elevator was observed to move from a position near neutral to near the full up position (full up just prior to impact), and the rudder was seen to move from a point near neutral to almost full left deflection (full left also just prior to impact). Once during the third turn, the aircraft momentarily rolled to a near wings-level attitude, but within one to two seconds had rolled back into a steep right turn. WRECKAGE AND IMPACT IMFORMATION The aircraft impacted flat grassy terrain about 200 feet northwest of runway 05/23, between the runway and the northwest parking ramp/taxiway. At the point of initial impact, it created a crater in the dirt about four feet wide, six feet long, and approximately three to four inches deep. It then bounced about 35 feet to the southeast, where it came to rest. The on-scene investigation showed that most of the damage to the aircraft occurred to the left wing and to the left side of the fuselage forward of the baggage area. The left main gear had been driven up through the left wing, and both the fore and aft wing spars had fractured near the gear attach point. The left wing spar-to-fuselage attach fitting had ripped lose from the fuselage, and the left side of the fuselage, from the instrument panel to the baggage area, had torn open. The left rail on the pilot's seat had torn lose from the structure, but the right rail remained attached. Much of the wooden primary fuselage structure on the left side of the cockpit area had fractured into small pieces, and the canopy was thrown free of the main wreckage. One blade of the propeller, which had a wood core and a composite cover, sheared off about 24 inches from the hub, and the other blade was undamaged. Pieces of the sheared blade were found in the initial impact crater. During the investigation the activation system for each flight controls was inspected for continuity and signs of possible binding or restriction of movement. The inspection of the elevator, rudder and flaps revealed no sign of pre-impact anomalies. However, an inspection of the aileron activation system revealed that in both wings, two three-wire electrical bundles were routed in a manner so as to be in contact with the aileron push-pull tube for almost their entire length (including where the tubes past through the small tube-routing holes in two of the inboard ribs). In the left wing, the bundles were routed under the aileron bellcrank mechanism, but in the right wing the bundles were routed through the center of the aileron bellcrank mechanism. In both wings, these wire bundles, which were determined to lead to the wingtip strobe and the wingtip smoke canister activation system, showed areas of rubbing and abrasion at numerous locations along their length . Also, there were numerous areas along the length of the aileron push-pull tubes where the anodizing had been rubbed off by contact with the sheath covering the wire bundles. In the right wing, the wire bundles were attached by tie-wraps to wing ribs both inboard and outboard of the bellcrank, but the looseness of the tie-wraps allowed both wires to move freely around the inside of the bellcrank support brackets. Both of the bundles had been abraded and dented where they passed through the bellcrank, and the sheath covering one of the bundles had been cut. In addition, the anodizing on the top surface of the lower bellcrank support bracket had been rubbed away, and portions of the bracket surface had been polished through repeated rubbing by the wire bundles. During the investigation, the aileron push-pull tube was moved through its full range of motion while moving the wire bundles in the right wing to different locations inside the bellcrank. While performing these tests, two situations were observed where the bundles became lodged in the bellcrank mechanism and caused binding in the aileron activation system. In both cases, the binding occurred while the push-pull tube was in a position where the right aileron was deflected in the up position (right roll). The first situation was when one of the bundles became lodged between the rod end assembly head on the outboard end of the push-pull tube and the wooden bellcrank bracket mounting block on the aft side of the main spar (see photograph #5). The second situation in which binding occurred was when one of the bundles became lodged between the tip of the bolt that passed through the aforementioned rod end assembly head and the upper surface of the lower bellcrank support bracket (see photograph #6). Because the two bundles were not tied together at the point where they passed through the bellcrank, it was possible to have both bundles lodged in separate locations at the same time. Also, as part of the investigation, the engine was partially disassembled and inspected, and no anomalies or evidence of any malfunction were found. AIRCRAFT INFORMATION The aircraft, which was manufactured in Bernay, France by Avions Mudrey & Cie, in 1975, was registered in the utility and acrobatic categories, and was purchased from another private party by its current owner about six or seven years prior to the accident. At the time of the accident, wingtip strobes and a wingtip smoke canister mounting and activation system were installed on the aircraft. According to Mudry Aviation Ltd., of Bunnell, Florida, the U.S. distributor of CAP aircraft, a CAP10B produced in 1975 would not have come from the factory with the wingtip strobes or smoke canister system installed. During the investigation, the mechanic who performed and signed off the last five annuals on the aircraft was interviewed. He said that he had installed the wingtip strobe system with the help of the owner, and that an FAA Form 337 had been completed for the installation. According to this mechanic, at the time of that installation, the wire bundles in both wings were routed so as to pass completely outside of the bellcrank mechanism. He also said that he did not participate in the installation of the smoke activation system, and felt quite sure that there was no entry in the aircraft logbook indicating that the installation had been made (the logbooks were never recovered). The mechanic also reported that the right wingtip strobe light had recently been inoperative, and it was his opinion that the pilot/owner had tried to troubleshoot that problem himself. The IIC was unable to determine who had installed the wingtip smoke system, or whether the pilot/owner had performed any work on the aircraft related to the problem with the wingtip strobe. TESTS AND RESEARCH After determining that it was possible for aileron jamming to occur due to the combined effect of the binding of the wire bundles passing through the right bellcrank and the friction caused by the wire bundles rubbing the push-pull tubes along most of their length, the IIC conducted flight tests in an equivalent CAP10B. During these tests, which were performed in order to determine if the aircraft was controllable with the ailerons stuck in the position were they jammed during the post-accident tests, the IIC was accompanied by CAP10B airshow pilot Montaine Mallet. To initiate the flight tests, the ailerons were placed in a deflected position approximating the position where they had jammed during the ground tests (about one-half to one inch of deflection at the inboard trailing edge). The tests were initiated both from a level flight attitude and during the pull-out of a vertical descent after a hammerhead stall maneuver. In both cases, the aircraft was not able to be rolled back to the wings-level position by use of left rudder deflection once it had passed through about 30 degrees of bank. In each case where the aircraft was allowed to pass 30 degrees of bank prior to attempting recovery with the rudder, it continued to roll to more than 70 degrees, and then started a slow descent while established in the steep bank. According to Ms. Mallet, the only possible means of recovery from this situation would be to allow the aircraft to continue rolling to the right, while attempting to gain altitude by alternately applying negative G loads (while the aircraft was inverted) and positive G loads (while the aircraft was upright). According to her, if the pilot were able to successfully perform this climbing maneuver, it may have been possible to gain enough altitude to bail out of the aircraft. During the test flight, it was also determined that to perform a left hammerhead turn similar to the one the pilot had performed just prior to reporting the locked control, the right aileron would be deflected to the full up position as the aircraft pivoted to the left at the top of the maneuver. ADDITIONAL INFORMATION As part of the investigation an autopsy was performed by The Forensic Institute at Holy Family Hospital, Spokane, Washington. The cause of death listed in the autopsy report was accidental, and was attributed to injuries related to blunt force impact. In addition, a forensic toxicology examination was performed by FAA's Toxicology and Accident Research Laboratory, and the findings were negative for carboxyhemoglobin, cyanide, ethanol, and screened drugs. The aircraft was released to Discount Aircraft Salvage, a representative of the owner, on October 24, 1996, at Deer Park, Washington.
loss of control due to a jammed aileron pushpull tube, due to improper installation/routing of electrical wire bundles through the right aileron bellcrank mechanism by unknown person(s).
Source: NTSB Aviation Accident Database
Aviation Accidents App
In-Depth Access to Aviation Accident Reports
