Session: 32-06 High Pressure Turbines 1

Paper Number: 129375

129375 - Adjoint Optimization of Non-Axisymmetric Endwall Contours to Reduce Losses in a High-Pressure Turbine 

As modern gas turbines continue to evolve, further efficiency gains become more difficult to realize. The parasitic loss from secondary flows become a larger component of the overall loss, making aerodynamic optimization of secondary effects ever more critical. Specifically, the secondary flow loss from the vortices created by the horseshoe vortex can become a dominant loss factor. There has been much research and development over the past 40 years into the mitigation of these secondary flow structures, but manufacturing advances are opening up the realm of what designs can be practically implemented.

      This paper discusses the development of an adjoint optimization method for three-dimensional endwall contouring using an adjoint solver in commercially available software. The primary goals of the study are to establish best practices for a holistic process of endwall optimization and compare the optimized design to the baseline geometry. An array of CFD analyses has been conducted for a high-pressure turbine stage to investigate the impact of the sensitivity radius parameter as well as to compare Reynolds’ Averaged Navier-Stokes turbulence models including k-epsilon, realizable k-epsilon, k-omega, and k-omega SST with regard to isentropic efficiency convergence. The results of these simulations are discussed and the effectiveness of the use of the adjoint model with respect to endwall contouring is discussed. Comparisons between the baseline geometry and optimized endwall shape are used to highlight how small changes in the endwall geometry impact the development of secondary flow structures such as the passage vortex.

 

Presenting Author: Austin Hendrickson The Ohio State University

Presenting Author Biography: Austin is a PhD student at the Ohio State University currently studying Aerospace Engineering. His primary interests include aerothermal studies in turbomachinery leveraging both experiment and CFD.

Authors:

Austin Hendrickson The Ohio State University
Spencer Sperling Honeywell Aerospace
Bao Nguyen Honeywell Aerospace
Richard Celestina Honeywell Aerospace
Jong Liu Honeywell Aerospace
Hakan Aksoy Honeywell Aerospace
Jeremy Nickol Honeywell Aerospace
Randall Mathison The Ohio State University