NTSB investigators either traveled in support of this investigation or conducted a significant amount of investigative work without any travel, and used data obtained from various sources to prepare this aircraft accident report.
The pilot was repositioning the helicopter from a rooftop helipad where it had just been refueled to a nearby airport. Video footage revealed that the helicopter lifted off of the helipad and simultaneously started to yaw to the left, consistent with a loss of tail rotor control. The helicopter completed one 360-degree rotation about the yaw (vertical) axis in a near level attitude while climbing. As it continued to rotate (spin) to the left, the helicopter deviated from a level attitude, pitching nose down and banking right, consistent with a loss of main rotor control. The helicopter moved away from the helipad, lost altitude, and impacted the street below. A postcrash fire erupted that consumed most of the fuselage and the forward section of the tailboom.
All major structural components of the helicopter were found at the accident site, and there was no evidence of an inflight failure of the airframe. Examination of the engine revealed that it was producing power at the time of impact. Further, the main rotor and tail rotor systems exhibited damage consistent with powered impact. Flight control continuity could not be confirmed due to fire and impact damage, and most components of the hydraulic system were severely fire damaged or destroyed preventing determination of their preimpact condition.
The NTSB determined that the loss of tail rotor and main rotor control resulted from a loss of hydraulic boost. This determination was made based on a series of deductions. A helicopter can enter a left yaw at takeoff for one of three reasons: 1) a loss of tail rotor effectiveness, 2) a loss of tail rotor drive, or 3) a loss of tail rotor control. In this accident, a loss of tail rotor effectiveness was unlikely because the reported wind speeds at nearby airports at the time of the accident were 4 knots or less. A loss of tail rotor drive was ruled out in this case based on the physical evidence indicating that the tail rotor was powered at ground impact. Thus, the left yaw at takeoff was likely due to a loss of tail rotor control.
A loss of tail rotor control can result from one of three circumstances: 1) a disconnect in the tail rotor pedal control system; 2) a restriction or jam in the pedal controls, or 3) a loss of hydraulic boost to the pedal controls. Although either a disconnect or a restriction/jam in the pedal controls would explain the helicopter's left yaw at takeoff, neither would explain the rapid loss of pitch and bank (main rotor) control that occurred after the first 360-degree yaw rotation that appears consistent with a loss of hydraulic boost to the main rotor controls. Therefore, a loss of hydraulic boost to the pedal controls, followed by a loss of hydraulic boost to the main rotor controls, most likely occurred.
The NTSB then evaluated scenarios that may have led to the complete loss of hydraulic boost to the main and tail rotor controls during takeoff. These scenarios include the following, with the last being the most likely, as described below:
Scenario 1 – Loss of hydraulic pressure due to mechanical failure and simultaneous failure of the yaw load compensator. In this scenario, a loss of pressure to the hydraulic system would result in the loss of hydraulic boost to the main and tail rotor servo controls. The yaw load compensator would provide partial hydraulic boost to the pedal controls unless the compensator failed, which would result in no hydraulic boost to the pedal controls. Because the yaw load compensator would likely have been functionally checked during the preflight hydraulics check, this scenario is considered unlikely because of the low probability of two separate failures occurring simultaneously.
Scenario 2 – A misconfiguration of the hydraulic system at the conclusion of the preflight hydraulic system checks. In this scenario, the pilot would have performed the preflight hydraulic system checks but would have failed to reset the "HYD TEST" button, the hydraulic cut-off switch, or both, at the end of the preflight hydraulic system checks. The accumulator check, which requires the pilot to activate and then reset the "HYD TEST" button, is the first of the two preflight hydraulic system checks. Activating the "HYD TEST" button depressurizes the main and tail rotor servo controls, depletes the yaw load compensator, but does not deplete the main rotor accumulators. The hydraulic cut-off test, which activates and then resets the hydraulic cut-off switch, is the second of the two preflight hydraulic system checks. Activating the hydraulic cut-off switch depressurizes the main and tail rotor servo controls, depletes the main rotor accumulators, but does not depressurize the yaw load compensator. The preflight hydraulic system checks require the pilot to visually confirm that the "HYD" warning light turns off after completion of each check, and this light would have remained illuminated had either or both the "HYD TEST" button and hydraulic cut-off switch not been reset by the pilot. Unless the pilot was performing the preflight hydraulic checks via tactile feel of the controls alone, without visual confirmation of the "HYD" warning light on the caution-warning panel, and did not verify all caution and warning lights were extinguished before takeoff, as required by the flight manual, the scenario of a misconfigured hydraulic system at the conclusion of the preflight hydraulic checks is unlikely.
Scenario 3 – Loss of pressure during the preflight hydraulic system accumulator check due to activation of the "HYD TEST" button combined with an unlocked collective stick. In this scenario, the pilot would have engaged the "HYD TEST" button and then moved the cyclic control stick to verify that the main rotor accumulators were functioning properly. If the collective stick was not locked during this check and one or more of the main rotor accumulators were depleted by the cyclic movements, the collective would have moved up rapidly. This uncommanded collective movement is caused by a design characteristic of the main rotor system in the AS350 helicopter. The uncommanded movement is prevented by engaging the collective lock as specified in the preflight checklist. Although accidents have occurred in which an unsecured collective stick moved up enough to cause an inadvertent liftoff (see NTSB accident investigations LAX01LA083 and LAX02TA299), postaccident ground testing with an exemplar helicopter showed that, at its estimated takeoff weight, the accident helicopter would not have become airborne or light on its skids due to uncommanded collective movement as a result of main rotor accumulator depletion alone.
Revision 3 of the AS350-B2 rotorcraft flight manual indicated that the preflight hydraulic system checks were to be conducted with the fuel flow control lever (FFCL) set between the "OFF" and "FLIGHT" detents. The pilot was trained in this procedure. During the ground tests, no heave (upward movement) was felt during the tests conducted with the FFCL set properly between the "OFF" and "FLIGHT" detents. However, the operator's checklist, which was likely used by the pilot, specified that the FFCL be set to the "FLIGHT" detent (a higher power setting) during the preflight hydraulic system checks. During the ground tests with the FFCL in the "FLIGHT" detent, when the collective moved up, a heave was felt by the occupants of the exemplar helicopter.
If the pilot did not lock the collective and performed the accumulator check with the FFCL in the "FLIGHT" detent per the operator's checklist, he may have been startled by an uncommanded increase in collective and the accompanying heave. The pilot may have reacted by manually increasing collective pitch, resulting in an unplanned takeoff. Once airborne, the lack of hydraulic boost to the pedals would have resulted in an uncontrolled left yaw, and, as all three main rotor accumulators became depleted, the main rotor controls would have lost hydraulic boost, resulting in a rapid loss of control. This scenario best matches the video evidence.
Because scenarios 1 and 2 are considered unlikely, scenario 3 is left as the most likely scenario for this accident. However, because of the damage to the hydraulic system components and because the helicopter was not equipped with any type of flight recording device, no determination could be made regarding the pilot's actions during performance of the preflight hydraulic checks or regarding the hydraulic system configuration when the helicopter became airborne. If a recorder system that captured cockpit audio, images, and parametric data had been installed, it would likely have enabled reconstruction of the sequence of events leading to the loss of control.