Lisa W. Griffin

Shape optimization of supersonic turbines using global approximation methods

There is growing interest to adopt supersonic turbines for rocket propulsion. However, this technology has not been actively investigated in the United States for the last three decades. To aid design improvement, a global optimization framework combining the radial-basis neural network (NN) and the polynomial response surface (RS) method is constructed for shape optimization of a two-stage supersonic turbine, involving O(10) design variables. The design of the experiment approach is adopted to reduce the data size needed by the optimization task. The combined NN and RS techniques are employed. A major merit of the RS approach is that it enables one to revise the design space to perform multiple optimization cycles. This benefit is realized when an optimal design approaches the boundary of a predefined design space. Furthermore, by inspecting the influence of each design variable, one can also gain insight into the existence of multiple design choices and select the optimum design based on other factors such as stress and materials consideration.

Simulations of the Unsteady Flow through the Fastrac Supersonic Turbine

Analysis of the unsteady aerodynamic environment in the Fastrac supersonic turbine is presented. Modal analysis of the turbine blades indicated possible resonance in crucial operating ranges of the turbopump. Unsteady computational fluid dynamics (CFD) analysis was conducted to support the aerodynamic and structural dynamic assessments of the turbine. Before beginning the analysis, two major problems with current unsteady analytical capabilities had to be addressed: modeling a straight centerline nozzle with the turbine blades and exit guide vanes (EGVs), and reducing run times significantly while maintaining physical accuracy. Modifications were made to the CFD code used in this study to allow the coupled nozzle/blade/EGV analysis and to incorporate Message Passing Interface (MPI) software. Because unsteadiness is a key issue for the Fastrac turbine [and future rocket engine turbines such as the Reusable Launch Vehicle (RLV)], calculations were performed for two nozzle-to-blade axial gaps. Calculations were also performed for the nozzle alone, and the results were imposed as an inlet boundary condition for a blade/EGV calculation for the large gap case. These results are compared to the nozzle/blade/EGV results.

Shape Optimization of Supersonic Turbines Using Response Surface and Neural Network Methods

Turbine performance directly affects engine specific impulse, thrust-to-weight ratio, and cost in a rocket propulsion system. A global optimization framework combining the radial basis neural network (RBNN) and the polynomial-based response surface method (RSM) is constructed for shape optimization of a supersonic turbine. Based on the optimized preliminary design, shape optimization is performed for the first vane and blade of a 2-stage supersonic turbine, involving O(10) design variables. The design of experiment approach is adopted to reduce the data size needed by the optimization task. It is demonstrated that a major merit of the global optimization approach is that it enables one to adaptively revise the design space to perform multiple optimization cycles. This benefit is realized when an optimal design approaches the boundary of a pre-defined design space. Furthermore, by inspecting the influence of each design variable, one can also gain insight into the existence of multiple design choices and select the optimum design based on other factors such as stress and materials considerations.

Neural Network and Response Surface Methodology for Rocket Engine Component Optimization

The goal of this work is to compare the performance of response surface methodology (RSM) and two types of neural networks (NN) to aid preliminary design of two rocket engine components. A data set of 45 training points and 20 test points, obtained from a semi-empirical model based on three design variables, is used for a shear coaxial injector element. Data for supersonic turbine design is based on six design variables, 76 training data and 18 test data obtained from simplified aerodynamic analysis. Several RS and NN are first constructed using the training data. The test data are then employed to select the best RS or NN. Quadratic and cubic response surfaces, radial basis neural network (RBNN) and back-propagation neural network (BPNN) are compared. Twolayered RBNN are generated using two different training algorithms, namely, solverbe and solverb. A two-layered BPNN is generated with Tan-Sigmoid transfer function. Various issues related to the training of the neural networks are addressed, including number of neurons, error goals, spread constants, and the accuracy of different models in representing the design space. A search for the optimum design is carried out using a standard, gradient-based optimization algorithm over the response surfaces represented by the polynomials and trained neural networks. Usually a cubic polynomial performs better than the quadratic polynomial but exceptions have been noticed. Among the NN choices, the RBNN designed using solverb yields more consistent performance for both engine components considered. The training of RBNN is easier as it requires linear regression. This coupled with the consistency in performance promise the possibility of it being used as an optimization strategy for engineering design problems.

A Study of the Effects of Tip Clearance in a Supersonic Turbine

Flow unsteadiness is a major factor in turbine performance and durability. This is especially true if the turbine is a high work design, compact, transonic, supersonic, counterrotating, or uses a dense drive gas. The vast majority of modern rocket turbine designs fall into these categories. In this study a parallelized unsteady three-dimensional Navier-Stokes analysis has been used to study the effects of tip clearance on the transient and time-averaged flow fields in a supersonic turbine. The predicted results indicate improved performance in the simulation including tip clearance. The main sources of the performance gains were: (1) a weakened shock system in the case with tip clearance, and (2) the fact that the reductions in the shock losses were greater than the losses introduced by tip clearance.

Detailed Aerodynamic Design Optimization of an RLV Turbine

A task was developed at NASA/Marshall Space Flight Center (MSFC) to improve turbine aerodynamic performance through the application of advanced design and analysis tools. There are four major objectives of this task: 1) to develop, enhance, and integrate advanced turbine aerodynamic design and analysis tools; 2) to develop the methodology for application of the analytical techniques; 3) to demonstrate the benefits of the advanced turbine design procedure through its application to a relevant turbine design point; and 4) to verify the optimized design and analysis with testing. The turbine chosen on which to demonstrate the procedure was a supersonic design suitable for a reusable launch vehicle (RLV). The hot gas path and blading were redesigned to obtain an increased efficiency over the baseline. Both preliminary and detailed designs were considered. The subject of the current paper is the optimization of the blading. To generate an optimum detailed design, computational fluid dynamics (CFD), response surface Copyright 2001 by the American Institute of Aeronautics and Astronautics, Inc. No copyright is asserted in the United States under Title 17, U.S. Code. The U.S. Government has a royalty-free liscense to exercise all rights under the copyright claimed herein for Governmental Purposes. All other rights are reserved by thecopyright owner methodology (RSM), and neural nets (NN) were used. The goal of the demonstration was to increase the total-tostatic efficiency, tit-s, of the turbine by eight points over the baseline design. The predicted rv-s of the optimized design was 10 points higher than the baseline.

Advancement of Turbine Aerodynamic Design Techniques

The Consortium for Computational Fluid Dynamics (CFD) Application in Propulsion Technology has been created at NASA/MSFC. Its purpose is to advance the state-of-the-art of CFD technology, to validate CFD codes and models, and to demonstrate the benefits attainable through the application of CFD in component design. Three teams are currently active within the Consortium: (1) the Turbine Technology Team, (2) the Pump Stage Technology Team, and (3) the Combustion Devices Technology Team. The goals, dynamics, and activities of the Turbine Team are the subjects of this paper.The Consortium is managed by NASA. The Turbine Team is co-coordinated by a NASA representative from the CFD area and an industry (Pratt & Whitney) representative from the area of turbine aerodynamic design. Membership of the Turbine Team includes experts in design, analysis, and testing from the government, industry, and academia. Each member brings a unique perspective, expertise, and experience to bear on the team’s goals of improving turbine efficiency and robustness while reducing the amount of developmental testing. To this end, an advanced turbine concept has been developed within the team using CFD as an integral part of the design process. This concept employs unconventionally high turning blades and is predicted to provide cost and performance benefits over traditional designs. This concept will be tested in the MSFC Turbine Airflow Facility to verify the design and to provide a unique set of data for CFD code validation. Currently, the team is developing and analyzing methods to reduce secondary and tip losses to further enhance turbine efficiency. The team has also targeted volute development as an area that could benefit from detailed CFD analysis.Copyright

Radial Turbine Preliminary Aerodynamic Design Optimization for Expander Cycle Liquid Rocket Engine

†† ‡‡ , A response surface-based dual-objective design optimization was conducted in the preliminary design of a compact radial turbine for an expander cycle rocket engine. The optimization objective was to increase the efficiency of the turbine while maintaining low turbine weight. Polynomial response surface approximations were used as surrogates, and the accuracies of such approximations improve by limiting the size of the domain and the number of variables for each response of interest. The optimization was accomplished in three stages using an approximate, one-dimensional model. In the first stage, a relatively small number of points were used to identify approximate constraint boundaries of the feasible domain and to reduce the number of variables used to approximate each one of the constraints. In the second stage, a moderate number of points in this approximate feasible domain were used to identify the region where both objectives had reasonable values. The last stage focused on obtaining high accuracy approximation in the region of interest with large number of points. The approximations were used to identify the Pareto front and to perform a global sensitivity analysis. Significant improvement was achieved compared to a baseline design.

Full and Partial Admission Performance of the Simplex Turbine

The turbines used in rocket-engine applications are often partial-admission turbines, meaning that the flow enters the rotor over only a portion of the annulus. These turbines have been traditionally analyzed, however, assuming full-admission characteristics. This assumption enables the simulation of only a portion of the 360-deg annulus with periodic boundary conditions applied in the circumferential direction. Whereas this traditional approach to simulating the flow in partial-admission turbines significantly reduces the computational requirements, the accuracy of the solutions has not been evaluated or compared to partial-admission data. In the current investigation, both full-admission and partial-admission three-dimensional unsteady Navier‐Stokes simulations were performed fo ra partial-admission turbine designed and tested at NASA Marshall Space Flight Center. The results indicate that the partial-admission nature of the turbine should be included in simulations to properly predict the performance and flow unsteadiness of the turbine.

Unsteady Flow in a Supersonic Turbine with Variable Specific Heats

Modern high-work turbines can be compact, transonic, supersonic, counter-rotating, or use a dense drive gas. The vast majority of modern rocket turbine designs fall into these Categories. These turbines usually have large temperature variations across a given stage, and are characterized by large amounts of flow unsteadiness. The flow unsteadiness can have a major impact on the turbine performance and durability. For example, the Space Transportation Main Engine (STME) fuel turbine, a high work, transonic design, was found to have an unsteady inter-row shock which reduced efficiency by 2 points and increased dynamic loading by 24 percent. The Revolutionary Reusable Technology Turbopump (RRTT), which uses full flow oxygen for its drive gas, was found to shed vortices with such energy as to raise serious blade durability concerns. In both cases, the sources of the problems were uncovered (before turbopump testing) with the application of validated, unsteady computational fluid dynamics (CFD) to the designs. In the case of the RRTT and the Alternate Turbopump Development (ATD) turbines, the unsteady CFD codes have been used not just to identify problems, but to guide designs which mitigate problems due to unsteadiness. Using unsteady flow analyses as a part of the design process has led to turbine designs with higher performance (which affects temperature and mass flow rate) and fewer dynamics problems. One of the many assumptions made during the design and analysis of supersonic turbine stages is that the values of the specific heats are constant. In some analyses the value is based on an average of the expected upstream and downstream temperatures. In stages where the temperature can vary by 300 to 500 K, however, the assumption of constant fluid properties may lead to erroneous performance and durability predictions. In this study the suitability of assuming constant specific heats has been investigated by performing three-dimensional unsteady Navier-Stokes simulations for a supersonic turbine stage.