Profsoyuznaya 65, Moscow, 117806, Russia

The object of the investigation is the task of nondestructive check of the transmission of turboprop aircraft engines (TPE) which transmits rotation of the free turbine (FT) to prop-fan. Because of big transmitted torque the elastic shaft is twisted. The angle of torsion depends on the torque's value. If we could estimate this angle in real-time this estimation can be used for the torque's identification. In turn the results of identification can be used in the engine control system for improving its characteristics. Moreover elastic stresses in the shaft’s material arising during its operating lead to the deterioration of the transmission characteristics and even can lead to the emergency situation when the shaft is broken. Current observations of the transmission's technical condition and taking measures which can prevent the risk of emergency situation is one of the most important problems of the intelligence adaptive system of the TPE condition monitoring. The aforementioned tasks are considered in the paper. On the example of the elastic transmission of turboprop aircraft engine with differential gearbox and two co-axial propellers the task of nondestructive check of long-length shaft transmitting the free turbine's torque to prop-fan is considered. The check task is removed to the identification of the shaft's torsion angle and to the fixation of the moment of material shaft plastic deformation beginning. The torsion angle information can be used for the turbine's torque identification in an effort to control and as a signal in system for leading cut-out of the engine in the case a dangerous situation preceding of the shaft's destroying. The influence of the two types of the noises (measurements’ noises and vibrational noises of the elastic shaft) on the accuracy of the shaft torsion angle estimation was investigated. Further it is very important to investigate the task of the reliability of the considered engine cut-out system from the viewpoint of the leading emergency system false action at rather big values of the estimation.