NTSB Factual Report

HISTORY OF FLIGHT

On May 11, 2005, at 0945 Pacific daylight time, a Robinson R44, N144SM, experienced a loss of directional control during takeoff, and impacted the runway coming to rest on its left side at Gustine Airport (3O1), Gustine, California. Helimark LLC, d.b.a. Silver State Helicopters, operated the helicopter under the provisions of 14 CFR Part 91 as an instructional flight. The helicopter sustained substantial damage. The certified flight instructor (CFI) and student pilot were not injured. Visual meteorological conditions prevailed for the local area instructional flight that departed Los Banos Municipal Airport (LSN), Los Banos, California, about 0930.

In a written statement to the National Transportation Safety Board, the CFI stated that both he and the student conducted the preflight inspection at LSN prior to departing for 3O1, and found no mechanical anomalies. They flew to 3O1, landed, and then performed "two normal patterns with normal approaches, still uneventful." After that, they performed two normal patterns with practice autorotations at the end of each pattern." At that point they decided to return to LSN for additional training.

The CFI reported that a normal takeoff profile was initiated from runway 36. About 8-10 feet above ground level, the CFI heard a clicking noise in the helicopter and decided to make a precautionary landing to find the source of the clicking noise. However, the helicopter began a gentle roll to the right. The CFI tried to apply left cyclic to stop the roll but realized he had no directional control of the helicopter. The helicopter veered off the right side of the runway, at a 10-degree bank angle. The right skid contacted the ground and the helicopter rolled over coming to rest on its left side. The engine continued to run after the helicopter came to a rest, so he pulled the mixture to shutdown the engine. Both the CFI and the student exited the helicopter through the pilot side door. Police and fire crews were alerted by airport personnel and arrived shortly thereafter to secure the scene. After the accident site was secured, the CFI located a bolt lying on the runway. He then inspected the helicopter and found that a bolt was missing from the connection between the left cyclic push-pull tube and the non-rotating swashplate.

An inspector from the Federal Aviation Administration (FAA) walked the runway area and found a nut that was consistent with the nut used in the connection between the left cyclic push-pull tube and the non-rotating swashplate.

AIRCRAFT INFORMATION

The helicopter was a Robinson R44, serial number 1098. A review of the helicopter's logbooks revealed the helicopter had a total airframe time of 1,601.8 hours at the last 100-hour inspection, completed on May 9, 2005. The accident occurred 2.6 hours after the helicopter had been signed off as airworthy.

The May 9, 2005, entry indicated in part: "installed new Bulkhead at top of mast fairing (p/n: C261-5 bulkhead), repaired bulkhead (p/n: C261-6 bulkhead)." According to Robinson Helicopter Company, the cyclic control rods are interconnected through the bulkheads. In order to remove and replace bulkheads, the cyclic control rods must first be detached from the swashplate.

The FAA's approved R44 Pilot's Operating Handbook requires the pilot to check the control rod ends for being "free without looseness" during the preflight inspection, and "verify that all fasteners are tight." It does not specify checking for the correct installation of the bolts that hold the control rods in place.

Tests and Research

An FAA inspector and a representative from Robinson Helicopter Company, a party to the investigation, inspected the airframe. The manufacturer's representative indicated that the nut and bolt found on the runway were the same type and part for the cyclic control input rod to the lower non-rotating swashplate assembly connection. The manufacturer's representative reported that a locking palnut completes the installation of the cyclic control input rods to the swashplate assembly. The FAA inspector indicated that during his search of the accident area he was not able to locate the locking palnut.

According to the FAA, parts located at the accident site were:

Bolt (NAS6605-19, found by CFI)

Nut (MS21042L5, found by FAA)

The bolt with what appeared to be a C141-1 spacer and a C115-1 spacer were found in place on the bolt shank, located 271 feet west and 20 feet north of the main wreckage. Not recovered at the accident site were the rod end, B330-16 Palnut, one C115-1 spacer, and AN960-516L Washer.

According to the manufacturer, the normal hardware required for the left swashplate ear connection are as follows:

NAS6605-19 or -20

C141-1 Spacer

C115-1 Spacer

D173-2 Rod End

C115-1 Spacer

Swashplate ear

AN960-516L Washer

MS21042L5 Nut (self-locking type)

B330-16 Palnut

The swashplate assembly, left push-pull tube, left rib (bulkhead), nut and bolt, and associated hardware were shipped to the National Transportation Safety Board metallurgical laboratory in Washington, D.C. for further examination.

The materials research engineer reported that examination of the bolt and nut revealed that the bolt was unbent, and that neither the bolt nor the nut showed any thread damage. The dimensions and characteristics of the bolt and nut matched the NAS6605-19 (bolt), and MS21042L5 Palnut (self-locking type) specified in manufacturer's Illustrated Parts Catalog (IPC) for the connection between the push-pull tube and the non-rotating swashplate. The IPC further identified an additional locking palnut to complete the installation.

The engineer noted that none of the holes in the lugs of the non-rotating swashplate exhibited any significant ovalization. When he examined the hole of the left lug of the non-rotating swashplate he noted some contact impressions within the hole on the bottom surface near the left side, which he characterized as a "circumferential ridge of raised metal." He was unable to place the shank portion of the bolt in the lughole; however, the threaded portion of the bolt was able to pass through the hole. The engineer noted that the bottom surface of the left lughole showed a number of deeper thread impressions near the middle of the hole and the ridge near the left side of the hole. The gouge on the right side of the ridge showed deformation of the material being pushed from the right to the left. The top surface of the left lughole showed less distinct thread impressions; however, he reported that the markings within the area of contact at the left edge indicated relative left-right sliding contact. In contrast, the right lug of the non-rotating swashplate had no deep thread impressions within the hole.

In addition, the engineer noted that the rotating swashplate fracture features were consistent with a bending, overstress separation and no pre-existing damage was noted.

ADDITIONAL INFORMATION

A similar accident occurred in Miami, Florida, on August 2, 2004, involving a Robinson R22 (NTSB number MIA04FA115). The investigation uncovered that the accident occurred shortly after maintenance that required disassembly of the bolted connection between the left cyclic push-pull tube and the non-rotating swashplate; the bolt for the connection was not in place. A metallurgical examination revealed thread imprints within the hole of the left lug of the non-rotating swashplate.

NTSB Probable Cause

the improper installation of the nut/bolt assembly connecting the left cyclic push-pull tube to the non-rotating swashplate by maintenance personnel, which led to a loss of roll control during takeoff.