Session: 06-03 Pressure Gain Combustion II

Paper Number: 126249

126249 - A Reduced Order Methodology for Optimizing Turbine Expanders Working With Rotating Detonation Combustors 

Redesigning a gas turbine cycle with creative concepts is one of the most critical possibilities for achieving a significant increase in efficiency. Pressure Gain Combustors (PGC) can be used in place of conventional combustors, since PGC contributes to a considerable gain in thermal efficiency while also emitting low NOx levels. Two of the most promising PGCs are the Pulse Detonation Combustor (PDC) in which a tube is cyclically filled with the fuel-oxidizer mixture that is then detonated, and Rotating Detonation Combustor (RDC) in which the combustion mixture is injected axially and the ignition is self-sustained by a rotating detonation wave. The PGC concept does, however, have technical challenges. A major one is the turbine integration to the outlet of the combustor due to the unsteady turbine inflow conditions. This unsteady exhausting flow causes turbomachinery components to operate under fluctuating off-design conditions. This fact results in reduced overall turbine performance. Toward addressing this issue, in this work, the turbine integration to the RDC and an optimization methodology for the turbine are discussed. Due to the highly fluctuating unsteady flow of the pressure gain combustors, the three-dimensional CFD simulation of turbine becomes very expensive. Thus, an alternative approach of using one-dimensional unsteady Euler model for the turbine is adopted. This unsteady one-dimensional Euler model uses source terms such as blade force in the momentum equation and shaft work in the energy equation. These source terms are computed from the steady-state turbine meanline analysis for several operating points and stored in a table. Depending on the current pressure at the blade leading edge, the 1D-Euler code interpolates the corresponding blade force and shaft work source term. The boundary conditions to the solver are the unsteady total pressure and total temperature at the inlet and constant static pressure at the outlet. The inlet boundary condition from the RDC exhaust is provided by an in-house 2D unsteady Euler model, which models a combustor with a single detonation wave rotating around the annulus. In this work, the High-Pressure Turbine (HPT) of NASA Energy Efficient Engine is used due to the available geometry and performance test results. An optimization methodology is carried out to minimize the entropy generation in turbine in unsteady operation. The optimization procedure uses Latin Hypercube Sampling (LHS) for generating random initial samples, Artificial Neural Network (ANN) as an approximation approach, and Complex Shape Method (CSM) as the searching algorithm. The optimization variable used here is the number of blades of all four blade rows representing the blade row solidities. For the optimization, 30 initial random samples were taken and entropy differences were calculated using the unsteady 1D Euler model. Then, an approximate solution was determined using ANN and new infill points are provided by CSM. Finally, the obtained optimized result is compared with the base design using both 1D unsteady Euler model and unsteady 3D-CFD simulation.

Presenting Author: Gokkul Raj Varatharajulu Purgunan Technical University of Berlin

Presenting Author Biography: I, Gokkul Raj, was born in Chennai, India. I am currently pursuing PhD in Technical University of Berlin. My work is with the full cycle simulation of rotating detonation combustor (RDC) with the turbomachinery upstream and downstream of the combustor. The paper which we are planned to submit for ASME 2024 is a novel method of how turbine can be integrated with RDC and also optimizing the turbine. In this work, we use reduced order model, unsteady 1D Euler solver and lower order optimizing methodology.

Authors:

Gokkul Raj Varatharajulu Purgunan Technical University of Berlin
Majid Asli Brandenburg University of Technology Cottbus-Senftenberg
Roman Klopsch Technical University of Berlin
Josh Meister Technical University of Berlin
Panagiotis Stathopoulos German Aerospace Center (DLR)