Propeller Synchrophase Angle Optimization of Turboprop-Driven Aircraft—An Experimental Investigation

Abstract

Propeller synchrophasing is an effective way of reducing interior noise and vibration of turboprop-driven aircraft. However, synchrophasing has achieved limited success in practice for the reason that the predetermined phase angles are not acoustically opti mized for maximum noise reduction during all flight conditions. An investigation has been conducted out which includes two folds: first, the noise vector based on laboratory experimental data has been modeled and second, optimal phase to acquire minimum noise is obtained via optimization search. An improved identification method of vector noise model which can be less dependent to noise phase message is presented. Compared with traditional methods, this method can greatly reduce the real-time requirement between phase optimization model and control model or sound acquiring model, so it can eliminate the influence which communication delay brings on identification precision. A synchrophasing experimental platform is established to verify the vector noise modeling. It adopts two propellers-driven servo motors to simulate the interior noise environment o f the aircraft. The influence of the date sampling condition on identification is also reseat ched. Ant colony optimization with two improvements is applied to phase optimiza tion o f four propellers. Simulation results show that the improved algorithm requires much less calculation.

Propeller Synchrophasing Control with a Cylindrical Scaled Fuselage Based on an Improved Data Selection Algorithm

“…Because the propellers are running at lower speed compared to that of turbofan, the low order harmonic components are in the dominating position in the spectrum analyses, which indicates that the high order harmonics can be neglected when the propeller signature model is identified. Huang et al [6] takes only the first six orders of the harmonic for the identification and good agreement is achieved between the real noise level and the noise prediction. To determine the accuracy of the model, the first ten harmonics are selected in this paper.…”

“…Obviously, the cost of the real flight is huge and the experiments are time-consuming, meaning they are not be suitable for further research on propeller synchrophasing control. Huang et al [6] and Sheng et al…”

“…Obviously, the cost of the real flight is huge and the experiments are time-consuming, meaning they are not be suitable for further research on propeller synchrophasing control. Huang et al [6] and Sheng et al[7] established a ground experimental platform to study the effectiveness of synchrophasing control and significant noise reduction was achieved. However, the existence of the fuselage is an inevitable factor that can affect the performance of the synchrophasing control, and is difficult to analyze due to the complex acoustic vibration couplings between the fuselage and the enclosed cabin space.…”

“…The measured sound pressure often fluctuates greatly due to the influence of flight speed, altitude, airflow variation, and circuit noise during the acquisition process. In order to reduce the impact of data fluctuations, the common solution is to collect noise information from more phase angles pairs than the required minimum pairs, thereby reducing the impact of unexpected data fluctuation on model accuracy [17,18].An algorithm for truncating the very beginning and tail of measured noise data is proposed by Huang et al [6] for minimizing transient data fluctuation caused by noise acquisition start and stop manipulation. The algorithm has obvious effects for some occasions, but the conclusion is made only based on experimental data and the length of the intercepted data lacks theoretical criteria.…”

“…An algorithm for truncating the very beginning and tail of measured noise data is proposed by Huang et al [6] for minimizing transient data fluctuation caused by noise acquisition start and stop manipulation. The algorithm has obvious effects for some occasions, but the conclusion is made only based on experimental data and the length of the intercepted data lacks theoretical criteria.…”