FAA-ASCENT project

 

FAA-ASCENT Project 75: Improved engine fan broadband prediction capabilities

Joint with United Technologies Research Center.

Project description:

The noise signature of contemporary turbofan engines is dominated by fan noise, both tonal and broadband. Accepted methods for predicting the tone noise have existed for many years.  As well, engine designers have methods for controlling or treating tonal noise. This is not the case for broadband noise. Thus, it is clear that further reductions in engine noise will require accurate prediction methods for the broadband noise to enable design decisions.  Interaction noise from the fan-stage is a dominant broadband mechanism in a modern high bypass engine and is created by the interaction of the turbulence in the fan-wakes with the fan exit guide vanes (FEGVs).  This project will leverage prior development of low-order models for the prediction of fan broadband interaction noise.  Gaps in the low-order approach will be addressed based on knowledge gained from computation and experimentation.

1st year 8/10/20-9/1/21. 2nd year 9/10/21-12/30/22, 3rd year 1/23 – 12/23

Thanks to strong teamwork, we have demonstrated we can machine learn rotor wake flow information.  RTRC researchers supplied the CFD results for 268 rotor cases. Undergraduate Berkely Watchmann devices methods for pre-processing all the data. MSSP student Zijie Huang developed machine learning algorithms for wake parameter prediction.  The following year end presentation highlights some of the findings: ASCENT presentation

Two conference papers explain much of the early advances. A new conference paper coming out in 2023 will explain the latest updates. The information is under the publications page.