Nateri K. Madavan

Multi-airfoil Navier-Stokes simulations of turbine rotor-stator interaction

An accurate numerical analysis of the flows associated with rotor-stator configurations in turbomachinery can be extremely helpful in optimizing the performance of turbomachinery. In this study the unsteady, thin-layer, Navier-Stokes equations in two spatial dimensions are solved on a system of patched and overlaid grids for a rotor-stator configuration from an axial turbine. The governing equations are solved using a finite-difference, upwind algorithm that is set in an iterative, implicit framework. Results in the form of pressure contours, time-averaged pressures, unsteady pressure amplitudes and phase are presented. The numerical results are compared with experimental data and the agreement is found to be good. The results are also compared with those of an earlier study which used only one rotor and one stator. The current study uses multiple rotors and stators and a pitch ratio that is much closer to the experimental ratio. Consequently the results of this study are found to be closer to the experimental data.

Application of Artificial Neural Networks to the Design of Turbomachinery Airfoils

The feasibility of applying artie cial neural networks to the aerodynamic design of turbomachinery airfoils is investigated. The design process involves dee ning a target pressure distribution, computing several e ows to adequately populate the design space in the vicinity of the target, training the neural network with this data, and, e nding a design that has a pressure distribution that is closest to the target. The last step is carried out using the network as a function evaluator. This design process is tested using an established e ow simulation procedure, a simple two-layer feedforward network and a conjugate gradient optimization technique. Results are presented for some validation tests as well as a complete design effort where the pressure distribution from a modern Pratt and Whitneyturbinewasused as a target. Theseresultsareveryencouraging and clearly warrant furtherdevelopment of the process for full three-dimensional design.

Improving the Unsteady Aerodynamic Performance of Transonic Turbines using Neural Networks

Summary A recently developed neural net-based aerody-namic design procedure is used in the redesign of atransonic turbine stage to improve its unsteadyaerodynamic performance. The redesign procedureused incorporates the advantages of both tradi-tional response surface methodology (RSM) andneural networks by employing a strategy calledparameter-based partitioning of the design space.Starting from the reference design, a sequence ofresponse surfaces based on both neural networksand polynomial fits are constructed to traverse thedesign space in search of an optimal solution thatexhibits improved unsteady performance. The pro-cedure combines the power of neural networks andthe economy of low-order polynomials (in terms ofnumber of simulations required and network train-ing requirements). A time-accurate, two-dimen-sional, Navier-Stokes solver is used to evaluate thevarious intermediate designs and provide inputs tothe optimization procedure. The procedure yieldeda modified design that improves the aerodynamicperformance through small changes to the refer-ence design geometry. These results demonstratethe capabilities of the neural net-based design pro-cedure, and also show the advantages of includinghigh-fidelity unsteady simulations that capture therelevant flow physics in the design optimizationprocess.A patent application that covers some of the original ideas inthis report has been filed by NASA. This report has been sub-mitted for review toward presentation at the 38th AIAA Aero-space Sciences Meeting and Exhibit, Jan. 10-13, 2000. Reno,NV.*Riverbend Design Services, Palm Beach Gardens. Florida.

Redesigning Gas-Generator Turbines for Improved Unsteady Aerodynamic Performance Using Neural Networks

A recently developed neural network-based aerodynamic design procedure is used in the redesign of a gasgenerator turbine stage to improve its unsteady aerodynamic performance. The redesign procedure used incorporatestheadvantagesofbothtraditionalresponse-surfacemethodologyandneuralnetworksbyemployingastrategy called parameter-based partitioning ofthedesign space.Starting from thereferencedesign, a sequenceofresponse surfaces based on both neural networks and polynomial e ts is constructed to traverse the design space in search of an optimal solution that exhibits improved unsteady performance. The procedure combines the power of neural networks and the economy of low-order polynomials (in terms of number of simulations required and network training requirements ). A time-accurate, two-dimensional, Navier ‐Stokes solver is used to evaluate the various intermediate designs and provide inputs to the optimization procedure. The procedure yields a modie ed design that improves theaerodynamicperformancethrough small changesto thereferencedesign geometry. Theseresults demonstrate the capabilities of the neural network-based design procedure and also show the advantages of including high-e delity unsteady simulationsthatcapturetherelevante owphysicsin thedesignoptimization process.

DNS of Flow in a Low-Pressure Turbine Cascade Using a Discontinuous-Galerkin Spectral-Element Method

A new computational capability under development for accurate and efficient high-fidelity direct numerical simulation (DNS) and large eddy simulation (LES) of turbomachinery is described. This capability is based on an entropy-stable Discontinuous-Galerkin spectral-element approach that extends to arbitrarily high orders of spatial and temporal accuracy and is implemented in a computationally efficient manner on a modern high performance computer architecture. A validation study using this method to perform DNS of flow in a low-pressure turbine airfoil cascade are presented. Preliminary results indicate that the method captures the main features of the flow. Discrepancies between the predicted results and the experiments are likely due to the effects of freestream turbulence not being included in the simulation and will be addressed in the final paper.

NASA Fixed Wing Project Propulsion Research and Technology Development Activities to Reduce Thrust Specific Energy Consumption

Abstract This paper presents an overview of the propulsion research and technology portfolio of NASA Fundamental Aeronautics Program Fixed Wing Project. The research is aimed at significantly reducing the thrust specific fuel/energy consumption of notional advanced fixed wing aircraft (by 60 percent relative to a baseline Boeing 737-800 aircraft with CFM56-7B engines) in the 2030 to 2035 time frame. The research investments described herein are aimed at improving propulsive efficiency through higher bypass ratio fans, improving thermal efficiency through compact high overall pressure ratio gas generators, and exploring the potential benefits of boundary layer ingestion propulsion and hybrid gas-electric propulsion concepts. 1.0 Introduction In 2008, the NASA Fixed Wing (FW) Project (then known as the Subsonic Fixed Wing project) funded system studies (Refs. 1 to 5) of notional future fixed-wing aircraft concepts that could be viable for commercial aviation in the 2030 to 2035 time frame. NASA required that these concept aircraft meet very aggressive metrics for fuel/energy consumption, emissions, and noise. A detailed description and assessment of propulsion systems and technologies considered in these system studies, and recommenda-tions for technologies that merited further investigation, were reported by Ashcroft, et al. Reference 6. The FW Project has identified several propulsion system themes that have emerged from the system studies as showing potential to significantly reduce fuel/energy consumption. These propulsion system themes are (1) cleaner, compact higher bypass ratio propulsion, (2) unconventional propulsion-airframe integration, and (3) hybrid gas-electric propulsion. The FW project has established the following high-level propulsion system technical challenges based on these themes to guide research investments; (1) compact high overall pressure ratio (50+ OPR) gas generators to increase fan bypass ratio and improve propulsive and thermal efficiency, (2) integrated boundary layer ingestion system to achieve a vehicle-level net system fuel/energy consumption benefit on a representative vehicle, by improving propulsive efficiency and reducing aircraft drag, and (3) high power density electric motors (four times existing state-of-the-art, SOA), and power management and distribution for future notional hybrid gas-electric propulsion systems. These investments are aimed at meeting the challenging metric of reducing thrust specific fuel/energy consumption by 60 percent relative to a baseline Boeing 737-800 aircraft with CFM56-7B engines.

Turbomachinery Airfoil Design Optimization Using Differential Evolution

An aerodynamic design optimization method that is based on an evolutionary algorithm known at Differential Evolution (DE) is described. DE is a simple, fast, and robust evolutionary strategy that has been proven effective in determining the global optimum for several difficult optimization problems. A Navier-Stokes solver is used to evaluate the various intermediate designs and provide inputs to the DE optimizer. The method is implemented on distributed parallel computers so that new designs can be obtained within reasonable turnaround times. A hybrid version of the method that uses the DE algorithm in conjunction with a neural network is also developed and shown to reduce overall computing time requirements. Results are presented for the inverse design of a turbine airfoil from a modern jet engine. The capability of the method to search large design spaces and obtain the optimal airfoils in an automatic fashion is demonstrated.

DNS of Low-Pressure Turbine Cascade Flows With Elevated Inflow Turbulence Using a Discontinuous-Galerkin Spectral-Element Method

Recent progress towards developing a new computational capability for accurate and efficient high–fidelity direct numerical simulation (DNS) and large–eddy simulation (LES) of turbomachinery is described. This capability is based on an entropy– stable Discontinuous–Galerkin spectral–element approach that extends to arbitrarily high orders of spatial and temporal accuracy, and is implemented in a computationally efficient manner on a modern high performance computer architecture. An inflow turbulence generation procedure based on a linear forcing approach has been incorporated in this framework and DNS conducted to study the effect of inflow turbulence on the suction– side separation bubble in low–pressure turbine (LPT) cascades. The T106 series of airfoil cascades in both lightly (T106A) and highly loaded (T106C) configurations at exit isentropic Reynolds numbers of 60,000 and 80,000, respectively, are considered. The numerical simulations are performed using 8th–order accurate spatial and 4th–order accurate temporal discretization. The changes in separation bubble topology due to elevated inflow turbulence is captured by the present method and the physical mechanisms leading to the changes are explained. The present results are in good agreement with prior numerical simulations but some expected discrepancies with the experimental data for the T106C case are noted and discussed.

Direct Numerical Simulation of the Turbulent Wake of an Infinitely Thin Flat Plate

A direct numerical simulation of the symmetric wake of an infinitely thin flat plate that is aligned parallel to a constant-pressure, uniform freestream is described. The boundary layers on both the upper and lower surfaces at the trailing edge of the plate are turbulent and statistically identical. The focus is on the near and intermediate wake regions. Results from a computation using a high-order-accurate, upwind-biased, finite-difference method with the geometry and flow conditions chosen to match an earlier experiment are presented. The development of the plate boundary layer and the evolution of the wake are documented. The streamwise evolution of the computed wake turbulence statistics is compared with experimental data. Overall agreement between the computations and the experiment is good. The computations are used to evaluate some theoretical results that have been suggested in the literature for the evolution of statistics along the centerline in the near-wake region. The computed results show that in the very near-wake region the centerline velocity varies as the square root of the distance from the trailing edge. This is in contrast to a cube-root variation that has been suggested in the literature based on theoretical considerations. Further downstream, the velocity exhibits a logarithmic recovery that is similar to several previous experiments.

Supercomputer applications in gas turbine flowfield simulation

The numerical simulation of the unsteady, three-dimen sional, viscous flow in a gas turbine stage is considered. Results from a three-dimensional, time-accurate Navier- Stokes simulation of rotor-stator interaction in an axial turbine stage are presented. The present study uses a fine grid in the spanwise direction to better resolve the complex three-dimensional flowfield, and complements earlier reported coarse-grid calculations. Several different features of the flowfield are analyzed and compared to earlier calculations and to experimental data wherever possible. Computer animation techniques are used to visualize various unsteady, three-dimensional features of the flow. The results demonstrate the capabilities of cur rent computing hardware in obtaining accurate simula tions of unsteady flows in turbomachines.