Primary finding
Probable cause
An in-flight separation of the helicopter’s horizontal stabilizer due to undetected fatigue cracking of the stabilizer spar, which resulted in a loss of control.
Investigator assessment
Analysis narrative
The commercial pilot was conducting a personal flight in the experimental, amateur-built helicopter. Several witnesses reported seeing the helicopter flying overhead. They reported that it appeared to by flying normally but that it then turned sideways, banked left, and descended to the ground. One witness reported hearing a breaking sound and then seeing the "back rotor" hanging from the helicopter. The horizontal stabilizer was found separated from the tailboom. Postaccident examination revealed that the horizontal stabilizer spar tube had fractured at the weld area just outboard of the mounting flange. Examinations of the fracture surfaces revealed features consistent with fatigue cracking that had initiated at multiple origins along the weld toe. Although no weld defects or corrosion were noted at the fatigue origins, large areas of both fracture faces were covered by red and brown corrosion products, indicating that the cracks were present and exposed for a considerable amount of time (at least many days but more likely many weeks). The fatigue origins were located on the aft surface of the spar and propagated generally forward. The origin location and direction of propagation were indicative of cyclic bending loads in the spar as if the tip of the stabilizer repetitively moved forward relative to the mount. The source of the cyclic bending loads was not clear but could have been the result of many different helicopter factors. These factors could have been unique to the accident helicopter or could be present on all similar helicopters. It is likely that the horizontal stabilizer separated in flight due to undetected fatigue cracking in the stabilizer spar, which resulted in the uncontrolled descent. As assembled, the fracture location and weld were partially hidden by the horizontal stabilizer's airfoil skin and not directly visible, which would have made any cracking difficult to see. Following the accident, the kit manufacturer issued a mandatory inspection and modification bulletin for the horizontal stabilizer, which detailed inspection criteria and spar replacement guidance if cracking was found.
Source record
Factual narrative
Horizontal Stabilizer The horizontal stabilizer and the tail rotor cross and sleeve assembly were sent to the NTSB Materials Laboratory in Washington, DC, for further examination. The phenolic sleeve on the tail rotor cross was fractured at the end flange. Examination of the flange fracture revealed a brittle overstress separation with no indications of discontinuities. The spar tube of the horizontal stabilizer was fractured just outboard of the mounting flange and adjacent to the inboard edge of the airfoil section. The forward side of the end plate was bent slightly outboard but no impact damage was apparent on either the end plate or the airfoil skin of the stabilizer. Figure 1: Photograph of fractured spar tube from horizontal stabilizer As manufactured, the mounting flange of the spar is welded (on the outboard side) to the spar tube and positioned immediately adjacent to the inboard edge of the airfoil skin. As assembled, the fracture location and weld are partially hidden by the airfoil skin and not directly visible. Most of the fracture followed the outboard edge of the mounting flange weld. As initially received, large areas of both fracture faces were obscured by red and brown corrosion products. These surfaces were cleaned for a more detailed examination. Magnified optical examinations of the fracture surfaces identified features consistent with fatigue cracking. Three fatigue origins were found on the outer surface of the spar tube at the outboard toe (edge) of the assembly weld. The fatigue propagated generally radially through the wall thickness with some circumferential spreading. The fatigue cracking progressed through about half of the total spar cross section. The fatigue origins and propagation were all on the aft portion of the spar. No corrosion or other obvious damage was apparent at any of the origins. Examinations also uncovered a rust-covered crack slightly inboard of and undercutting the main fracture plain. The undercutting crack intersected the outboard fatigue and formed part of the overall fracture. Magnified examinations of the crack faces without opening revealed fatigue features with at least one fatigue origin on the outer surface of the spar. The crack morphology was consistent with additional fatigue origins in the unopened portion of the crack. The remaining fracture surface displayed separation features and deformation patterns consistent with overstress separation. The deformation pattern was consistent with forces associated with the stabilizer tip moving forward and outboard. Visually, the weld exhibited good workmanship with no apparent undercutting, weld cracking, or surface discontinuities. A small pore and a single area of lack of fusion were uncovered by the weld fracturing in the overstress regions. The inboard fracture face was viewed using a scanning electron microscope (SEM) after it was cleaned and the corrosion removed. The fracture area near one fatigue origin point displayed a corrosion-damaged surface, while fracture two other surfaces were much less damaged by corrosion and displayed features consistent with fatigue cracking in alloy steels. Energy dispersive x-ray spectra of the spar material acquired during SEM examinations were typical of an AISI2 4100 series alloy steel as indicated on the stabilizer engineering drawing. Ignition System Examinations of the LSE Plasma III CD Ignition Module (S/N: 43546), the Hall Effect Module (S/N: 1440), and the ignition coils (P/N: 356120) were conducted on March 31, 2015, at the facilities of Light Speed Engineering in Santa Paula, California, with NTSB oversight. No evidence of preimpact mechanical malfunction was noted during the examinations. Handheld GPS Device A Garmin GPSMAP 296 hand-held GPS device was retrieved from the accident site and sent to the NTSB Recorders Laboratory for examination. No tracklog information was present on the device after download using the manufacturer procedures. As a result of the fatigue cracking of the horizontal stabilizer spar found during this investigation, Safari Helicopter issued a mandatory inspection and modification bulletin titled, "Horizontal Stabilizer Mandatory Inspection and Modification," on April 6, 2015. The bulletin specified a liquid penetrant inspection (LPI) of the horizontal stabilizer as follows: "On helicopters in operation less than 24 months, this inspection should be accomplished at the next annual condition or 100-hour inspection. On helicopters older than 24 months, the inspection should be accomplished before next flight. This inspection should be added to the annual condition inspection for your particular aircraft." The bulletin stated that, if cracking was found on the horizontal stabilizer spar, it must be replaced with a new spar. If no cracking was found during LPI inspection, Safari indicated that a steel tube insert must be installed into the end of the stabilizer to extend the "full length of the tail boom mount." The bulletin also noted that the loss of the horizontal stabilizer "would change the attitude of the helicopter." Given that the stabilizer provides a downward-acting force on the tail section of the helicopter, this attitude change would be experienced by the pilot as an abrupt, uncommanded nose-down pitch. An autopsy of the pilot was conducted under the authority of Montgomery County Forensic Services, Conroe, Texas. The cause of death for the pilot was attributed to "multiple blunt injuries." The FAA Bioaeronautical Sciences Research Laboratory, Oklahoma City, Oklahoma, performed toxicology testing for the pilot. The pilot's toxicology results were negative for carbon monoxide and alcohol. The medications colchicine, doxazosin, and valsartan were detected. The pilot reported use of doxasozin and valsartan on his most recent FAA medical examination. Both of these prescription medications are used to treat high blood pressure and are not considered to pose a hazard to flight safety. The prescription medication colchicine, which is used to prevent and/or relieve the pain from gout attacks in adults, was not reported previously to the FAA. Some possible side effects of colchicine include nausea, diarrhea, stomach cramps, and weakness. The NTSB investigator-in-charge (IIC) did not travel to the accident scene. According to photographs taken by the FAA on-site, the main wreckage consisting of the fuselage, main rotor, tail and tail rotor came to rest at the fenceline of a treed area. The horizontal stabilizer had separated from the tail boom and was found about 65 feet northeast of the fuselage. The right door frame was found about 85 feet northeast of the fuselage. The helicopter wreckage was examined at the facilities of Air Salvage of Dallas on February 24-25, 2015, under the supervision of the NTSB IIC. The examination revealed that the horizontal stabilizer's spar (a steel tube) had separated at the weld area just outboard of a round steel plate (the stabilizer mounting flange) that was welded to the spar. The fracture exhibited signatures of corrosion and fatigue at the weld line. The inboard portion of the spar remained bolted to the tail boom at the mounting flange. Paint transfers were observed on one tail rotor blade. These transfers were located on the side of the blade where the horizontal stabilizer would be mounted. Additionally, both tail rotor blades exhibited impact damage on the leading and trailing edges of the blades. Flight control continuity was established from the cockpit flight controls through the tail rotor system. The control pedal cable leading from the tail rotor assembly to the right control pedal was found in the full right control pedal position. (In this helicopter, the right pedal is used to counter the torque of the main rotor.) The stationary swash plate (controlling right/left, fore, and aft cyclic) was fractured and separated at all three push-pull tub