Engineering topics
Publications and source records attributed to Keith, Theo G., Jr..
Aeroelastic codes with advanced capabilities for modeling flow require substantial computational time. On the other hand, fast-running linear aeroelastic codes lack the capability to model three-dimensional, transonic, vortical, and viscous flows. The goal of this work was to develop an aeroelastic code with accurate modeling capabilities and small computational requirements.
Centrifugal compressors are very widely used in the turbomachine industry where low mass flow rates are required. Gas turbine engines for tanks, rotorcraft and small jets rely extensively on centrifugal compressors for rugged and compact design. These compressors experience problems related with unsteadiness of flowfields, such as stall flutter, separation at the trailing edge over diffuser guide vanes, tip vortex unsteadiness, etc., leading to rotating stall and surge. Considerable interest exists in small gas turbine engine manufacturers to understand and eventually eliminate the problems related to centrifugal compressors. The geometric complexity of centrifugal compressor blades and the twisting of the blade passages makes the linear methods inapplicable. Advanced computational fluid dynamics (CFD) methods are needed for accurate unsteady aerodynamic and aeroelastic analysis of centrifugal compressors. Most of the current day industrial turbomachines and small aircraft engines are designed with a centrifugal compressor. With such a large customer base and NASA Glenn Research Center being, the lead center for turbomachines, it is important that adequate emphasis be placed on this area as well. Currently, this activity is not supported under any project at NASA Glenn.
Under the Base R&T and Ultra Efficient Engine Technology programs, the NASA-Goddard Space Flight Center Compressor Branch is investigating flow control strategies required to increase the loading and efficiency of core compressors while maintaining current levels of operability. Flow-control strategies being studied include advanced casing treatments, wall jet injection, and blade-tip injection for compressor stability enhancement, directed jets for surface boundary layer control, and vortex-generating devices. The use of computational fluid dynamics (CFD) simulations to assess the effectiveness of flow-control devices and to guide their design is a key element in this research. CFD simulations serve to screen potential flow-control concepts at a lower cost than executing physical experiments in turbomachinery facilities. CFD simulations also provide guidance in designing physical experiments for those flow control concepts, which appear promising.
The Institute for Computational Mechanics in Propulsion (ICOMP) was formed to develop techniques to improve problem-solving capabilities in all aspects of computational mechanics related to propulsion. ICOMP is operated by the Ohio Aerospace Institute (OAI) and funded via numerous cooperative agreements by the NASA Glenn Research Center in Cleveland, Ohio. This report describes the activities at ICOMP during 1999, the Institute's fourteenth year of operation.
The objective of this research is to develop an efficient numerical algorithm with unstructured grids for the computation of three-dimensional chemical reacting flows that are known to occur in combustion components of propulsion systems. During the grant period (1996 to 1999), two companion codes have been developed and various numerical and physical models were implemented into the two codes.
It has been estimated that the noise levels in aircraft engine transmissions can be reduced by as much as 10 dB through the use of journal bearings. The potential benefits of lower noise levels include reduced wear, longer gear life and enhanced comfort for passengers and crew. Based on this concept the journal-thrust wave bearing was analyzed and its performance was evaluated. Numerical codes, developed over the past 30 years by Dr. Dimofte, were used to predict the performance of the bearing. The wave bearing is a fluid film bearing and therefore was analyzed using the Reynolds pressure equation. The formulation includes turbulent flow concepts and possesses a viscosity-temperature correction. The centrifugal growth of the bearing diameter and the deformation of the bearing under gear loads were also incorporated into the code. An experimental rig was developed to test the journal-thrust wave bearing.
A summary of the work performed under NASA grant NCC3-605 is presented. More details can be found in the cited references. This grant led to the development of relatively faster aeroelastic analyses methods for predicting flutter and forced response in fans, compressors, and turbines using computational fluid dynamic (CFD) methods.
The propulsion element of the NASA Advanced Subsonic Technology (AST) initiative is directed towards increasing the overall efficiency of current aircraft engines. This effort requires an increase in the efficiency of various components, such as fans, compressors, turbines etc. Improvement in engine efficiency can be accomplished through the use of lighter materials, larger diameter fans and/or higher-pressure ratio compressors. However, each of these has the potential to result in aeroelastic problems such as flutter or forced response. To address the aeroelastic problems, the Structural Dynamics Branch of NASA Glenn has been involved in the development of numerical capabilities for analyzing the aeroelastic stability characteristics and forced response of wide chord fans, multi-stage compressors and turbines. In order to design an engine to safely perform a set of desired tasks, accurate information of the stresses on the blade during the entire cycle of blade motion is required. This requirement in turn demands that accurate knowledge of steady and unsteady blade loading is available. To obtain the steady and unsteady aerodynamic forces for the complex flows around the engine components, for the flow regimes encountered by the rotor, an advanced compressible Navier-Stokes solver is required. A finite volume based Navier-Stokes solver has been developed at Mississippi State University (MSU) for solving the flow field around multistage rotors. The focus of the current research effort, under NASA Cooperative Agreement NCC3- 596 was on developing an aeroelastic analysis code (entitled TURBO-AE) based on the Navier-Stokes solver developed by MSU. The TURBO-AE code has been developed for flutter analysis of turbomachine components and delivered to NASA and its industry partners. The code has been verified. validated and is being applied by NASA Glenn and by aircraft engine manufacturers to analyze the aeroelastic stability characteristics of modem fans, compressors and turbines.
Under this grant, Version 4 of the three-dimensional Navier-Stokes aeroelastic code (TURBO-AE) has been developed and verified. The TURBO-AE Version 4 aeroelastic code allows flutter calculations for a fan, compressor, or turbine blade row. This code models a vibrating three-dimensional bladed disk configuration and the associated unsteady flow (including shocks, and viscous effects) to calculate the aeroelastic instability using a work-per-cycle approach. Phase-lagged (time-shift) periodic boundary conditions are used to model the phase lag between adjacent vibrating blades. The direct-store approach is used for this purpose to reduce the computational domain to a single interblade passage. A disk storage option, implemented using direct access files, is available to reduce the large memory requirements of the direct-store approach. Other researchers have implemented 3D inlet/exit boundary conditions based on eigen-analysis. Appendix A: Aeroelastic calculations based on three-dimensional euler analysis. Appendix B: Unsteady aerodynamic modeling of blade vibration using the turbo-V3.1 code.
An oil-lubricated wave journal-thrust bearing assembly was successfully tested at conditions found in general aviation engine gearboxes. The bearing performed well at both steady state conditions and in start-stop tests. It ran stably under all loading conditions, including zero load, at all speeds up to 16 000 rpm. The bearing carried 25 percent more load than required for the gearbox application, supporting 8900 N (94 bars average pressure), and showed very good thermal stability. 450 start-stop cycles were also performed, including 350 cycles without oil supply during starting and stopping. Test results and numerical predictions were in good agreement.
The Institute for Computational Mechanics in Propulsion (ICOMP) was formed to develop techniques to improve problem-solving capabilities in all aspects of computational mechanics related to propulsion. ICOMP is operated by the Ohio Aerospace Institute (OAI) and funded via numerous cooperative agreements by the NASA Glenn Research Center in Cleveland, Ohio. This report describes the activities at ICOMP during 1998, the Institutes thirteenth year of operation.
A gas journal bearing, with a wavy surfaces was tested in a range of speeds up to 18,000 RPM to determine its stability in an unloaded condition as a function of the wave amplitude. The bearing, was 50 mm in diameter, 58 mm long and had 0.01 65 mm radial clearance. Three waves were created on the inner surface by deforming the bearing sleeve. The ratio of the wave amplitude to the radial clearance (the wave amplitude ratio) was varied from zero to 0.3.
The friction and wear properties of two groups of magnetic fluids, one developed at NASA Lewis Research Center and a commercial fluid, were evaluated for boundary lubrication. Friction and wear measurements were made using a pin-on-disk apparatus. Three different ball materials were evaluated, (1) 440C, (2) Al2O3, and (3) Si3N4 against 440C disks. The first class of magnetic fluids have a low vapor pressure hydrocarbon base oil and are suitable for space application. Four variations of this fluid were evaluated: (1) the base oil, (2) base oil with anti-wear additives, (3) a 100 Gauss strength magnetic fluid, and (4) a 400 gauss magnetic fluid. The commercial fluid base oil and four different magnetic particle concentration levels have been evaluated. A space qualified fluorinated lubricant was tested for base line comparison. Hardness, optical microscopy, surface profilometry, and surface analysis were used to characterize the test specimens. Friction was unaffected by the concentration of magnetic particles. Wear rates for magnetic fluids were slightly higher than the base oil. The low vapor pressure magnetic fluid has better wear characteristics than the space qualified fluorinated lubricant.
The National Aeronautics and Space Administration (NASA) sent a mission to the martian surface, called Mars Pathfinder. The mission payload consisted of a lander and a rover. The primary purpose of the mission was demonstrating a novel entry, descent, and landing method that included a heat shield, a parachute, rockets, and a cocoon of giant air bags. Once on the surface, the spacecraft returned temperature measurements near the Martian surface, atmosphere pressure, wind speed measurements, and images from the lander and rover. The rover obtained 16 elemental measurements of rocks and soils, performed soil-mechanics, atmospheric sedimentation measurements, and soil abrasiveness measurements.
Over the past several years, numerical and experimental investigations have been performed on a waved journal bearing. The research work was undertaken by Dr. Florin Dimofte, a Senior Research Associate in the Mechanical Engineering Department at the University of Toledo. Dr. Theo Keith, Distinguished University Professor in the Mechanical Engineering Department was the Technical Coordinator of the project. The wave journal bearing is a bearing with a slight but precise variation in its circular profile such that a waved profile is circumscribed on the inner bearing diameter. The profile has a wave amplitude that is equal to a fraction of the bearing clearance. Prior to this period of research on the wave bearing, computer codes were written and an experimental facility was established. During this period of research considerable effort was directed towards the study of the bearing's stability. The previously developed computer codes and the experimental facility were of critical importance in performing this stability research. A collection of papers and reports were written to describe the results of this work. The attached captures that effort and represents the research output during the grant period.
A pre-release version of the Navier-Stokes solver (TURBO) was obtained from MSU. Along with Dr. Milind Bakhle of the University of Toledo, subroutines for aeroelastic analysis were developed and added to the TURBO code to develop versions 1 and 2 of the TURBO-AE code. For specified mode shape, frequency and inter-blade phase angle the code calculates the work done by the fluid on the rotor for a prescribed sinusoidal motion. Positive work on the rotor indicates instability of the rotor. The version 1 of the code calculates the work for in-phase blade motions only. In version 2 of the code, the capability for analyzing all possible inter-blade phase angles, was added. The version 2 of TURBO-AE code was validated and delivered to NASA and the industry partners of the AST project. The capabilities and the features of the code are summarized in Refs. [1] & [2]. To release the version 2 of TURBO-AE, a workshop was organized at NASA Lewis, by Dr. Srivastava and Dr. M. A. Bakhle, both of the University of Toledo, in October of 1996 for the industry partners of NASA Lewis. The workshop provided the potential users of TURBO-AE, all the relevant information required in preparing the input data, executing the code, interpreting the results and bench marking the code on their computer systems. After the code was delivered to the industry partners, user support was also provided. A new version of the Navier-Stokes solver (TURBO) was later released by MSU. This version had significant changes and upgrades over the previous version. This new version was merged with the TURBO-AE code. Also, new boundary conditions for 3-D unsteady non-reflecting boundaries, were developed by researchers from UTRC, Ref. [3]. Time was spent on understanding, familiarizing, executing and implementing the new boundary conditions into the TURBO-AE code. Work was started on the phase lagged (time-shifted) boundary condition version (version 4) of the code. This will allow the users to calculate non-zero interblade phase angles using, only one blade passage for analysis.
Sonoluminescence is the effect of producing light from sound and occurs when a gas bubble is trapped in a fluid filled cavity and is forced to collapse under a barrage of sound waves. Frenzel and Schultes discovered this phenomenon in 1934 while exposing acoustic waves to photographic plates. This effect was not well understood until 1988 when Crum and Gaitan discovered the necessary conditions for producing single bubble sonoluminescence in the laboratory. The luminescence is a result of the bubble violently collapsing from sound waves and this shares a close association with vibratory cavitation. Cavitation erosion is known to cause damage to rotational machinery when the collapse is near to surfaces due to the high pressures associated with bubble collapse. With these high pressures and temperatures there is a considerable amount of damage to the outside layer of a bearing, thereby, reducing its useful life. An experiment was constructed to generate sonoluminescence in the laboratory in order to obtain a greater understanding of this phenomenon and its association with bubble cavitation. Most of the research was done to investigate how to obtain single bubble sonoluminescence under different conditions and to determine how to detect it. Success in this has inspired several theories on how to use the methods for generating sonoluminescence to control cavitation in fluids under industrial conditions.
Under this grant, the flutter versions of the three-dimensional Navier-Stokes aeroelastic code (TURBO-AE) have been developed. The TURBO-AE aeroelastic code allows flutter calculations for a fan, compressor, or turbine blade row. This code models a vibrating three-dimensional bladed disk geometry and the associated unsteady flow (including shocks, and viscous effects) to calculate the aeroelastic instability using a work-per-cycle or eigenvalue approach. Steady and unsteady calculations have been performed with the 3D Navier-Stokes (TURBO) code to gain familiarity with the code. Calculations were done with the NASA Rotor-67 configuration which has been used as a standard test case by turbomachinery researchers. Additional calculations have been performed for the NASA/GE EEE fan and a proprietary configuration.