Effect of Gradient Reconstruction Method on Transonic Fan Performance
Conference Paper
Overview
Identity
Other
View All
Overview
abstract
2017 by Forrest L. Carpenter. This paper presents an investigation of the choice of gradient evaluation method on the performance of a transonic fan, namely NASA Rotor 67. Turbulent solutions were computed using gradients found via the Green-Gauss, un-weighted least-squares with QR decomposition (LSQR), weighted LSQR (WLSQR), and weighted essential non-oscillatory (WENO) methods. Partial speedlines at the design wheel speed were computed for each gradient case, and the flow at the near peak efficient point was analyzed. The un-weighted LSQR gradient solutions were found to under-predict all performance quantities, while the other three methods showed good agreement with experiment. Increased flow separation was found to be associated with the un-weighed LSQR solutions, leading to higher losses.