COMPUTATIONAL PREDICTION OF A LARGE-SCALE HP TURBINE FLOW AGAINST MEASURED AERODYNAMIC DATA

Open Access
Author:
Doshi, Mitansh Sharad
Graduate Program:
Aerospace Engineering
Degree:
Master of Science
Document Type:
Master Thesis
Date of Defense:
November 26, 2018
Committee Members:
  • Dr. Cengiz Camci, Thesis Advisor
Keywords:
  • Turbine
  • CFD
  • Validation
  • NGV
  • Rotor
  • AFTRF
  • Turbulence model
  • K-Omega
  • Efficiency
Abstract:
This thesis presents a Reynolds-averaged Navier Stokes (RANS) equations-based computational validation of the Axial Flow Turbine Research Facility (AFTRF). The research turbine design was based on NASA’s E3 "Energy Efficient Engine" concept with 23 stationary nozzle guide vanes and a 29 blade high pressure (HP) turbine rotor. This large-scale and low-speed turbomachinery research facility provides high-resolution aerodynamic measurements from the turbine stage for the assessment of computational simulations. The finite volume-based general purpose fluid dynamics solver, Star CCM+, coupled with the k-ω SST turbulence model and the "Gamma transition" flow model were used. Various performance parameters were measured, including velocity profiles, nozzle/blade airfoil static pressure coefficients, and total pressure. The previously measured experimental data sets and boundary conditions from the AFTRF were used in a computational validation. NGV (Nozzle Guide Vanes) and rotor validations were performed and total-to-total efficiency was discussed. The present computational effort uses a "mixing plane" based stationary to rotating interface for stage calculations. A grid dependency assessment has been performed both on NGV and rotor flows. The computational results obtained at NGV-intraspace and rotor exit are compared to five-hole-probe based experimental data. The stage exit data from a Kiel probe are also compared to the current simulations. The current study concludes that the present computational model effectively predicts AFTRF aerodynamic flow features with good spatial resolution. An attempt is also made to compare the total-to-total efficiency distribution in the spanwise direction. The study concludes that this computational approach can be effectively used in turbine secondary flow reduction, tip leakage flow mitigation, unsteady flow computations and finally energy efficiency improvements.