Affordable Access

On the use of CFD for improvement of Low-Fidelity Blade-Vortex Interaction Prediction Code

Authors
  • Reboul, Gabriel
  • Falissard, Fabrice
Publication Date
May 13, 2019
Source
HAL-SHS
Keywords
Language
English
License
Unknown
External links

Abstract

High-fidelity simulation performed with the CFD code elsA has been used to improve a fast prediction code, named ARHIS, dedicated to the prediction of the aerodynamic response of a blade in interaction with a vortex. To do so, comparisons have been made on a well documented blade-vortex interaction configuration. After validation of the computation methodology against available measurements, discrepancies between elsA and ARHIS solutions have been pointed out at the blade tip, due to the bidimensional approach used in the latter. Consequently, a new model of tip loss factor has been developed in ARHIS providing a better agreement between fast and high fidelity simulations. The maximal error in terms of amplitude of the lift jump due to the blade vortex interaction goes from 20,1% to 5,2%. This new model has been applied to several cases to assess the impact of the blade shape by considering modification of the aspect ratio, sweep angles, parabolic tip and tapered blade. Even if not designed specifically to account for modification of the blade shape, it is clear that the new tip loss factor model still improves the agreement between AHRIS and elsA results. Finally, the improved model has been tested for the evaluation of the noise emission of complete rotors in descent flight including a straight and a double swept rotor blade. In all cases, the new model induces a reduction of the predicted noise level of about 0,5 to 1 dB. This noise reduction appears to be an improvement in one case but a deterioration in the other one maybe because of a compensation of errors in the entire computational chain. This new modeling needs consequently to be applied on a larger database to conclude on his efficiency in terms of noise predictions.

Report this publication

Statistics

Seen <100 times