This is an excerpt from the AAIB Bulletin that Nr Fairy mentioned.
Gazelle yaw characteristics
In the majority of civil and military cases, loss of yaw control occurred in the hover or at low forward speed in light winds from the right. A few occurred in stronger winds or with wind from the left.
Both inexperienced and highly experienced pilots were involved in the military accidents and loss of control of pitch and roll during the subsequent high rates of rotation was a commonfeature. An 'optimised fenestron' was fitted to military Gazelles in the early 1980s as part of a weight upgrade programme.
The optimised fenestron had revised duct and hub fairings but did not appear to improve the incidence of sudden loss of yaw control.
The sudden loss of yaw control was attributed to 'fenestron stall'and, in response to concern, the Ministry of Defence (MOD) sponsoreda trial by the manufacturer, Eurocopter France, to investigate the phenomenum.
The trial took place in 1992/93 and demonstrated that, in conditions of low natural wind, a relatively small left pedal input of 5% (of total pedal travel) from the hover position can result in a yaw rate of 150°/sec being achieved in 10seconds.
It also showed that high yaw rates to the left (165°/sec)can be rapidly arrested by application of full right pedal without any tendency for aerodynamic stall of the fenestron.
The MOD advice included a statement that the extremely rapid build up of yaw rate in these circumstances was exacerbated if the SASwas not engaged.
The MOD trial did not establish why a small pedal input can result in the rapid build up of very high yaw rates. However, an earlier study, in 1991, by Westland Helicopters Limited had suggested that the trigger mechanism might involve a coupling of fenestron rotor induced swirl with the circulation contained in the main rotor tip vortices which may become aligned with the fenestron in certain flight conditions.
The study also suggested that considerationshould be given to changing the direction of rotation of the fenestron to become top-blade-aft which would probably solve the interactional aerodynamic problem.
Subsequent fenestron-equipped helicopters such as the SA365 Dauphin, EC135 and EC120 have top-blade-aft fenestron rotation; they are not known to suffer from sudden lossof yaw control.