Michael G. Dunn

Convective Heat Transfer and Aerodynamics in Axial Flow Turbines

The primary focus of this paper is convective heat transfer in axial flow turbines. Research activity involving heat transfer generally separates into two related areas: predictions and measurements. The problems associated with predicting heat transfer are coupled with turbine aerodynamics because proper prediction of vane and blade surface-pressure distribution is essential for predicting the corresponding heat transfer distribution. The experimental community has advanced to the point where time-averaged and time-resolved three-dimensional heat transfer data for the vanes and blades are obtained routinely by those operating full-stage rotating turbines. However, there are relatively few CFD codes capable of generating three-dimensional predictions of the heat transfer distribution, and where these codes have been applied the results suggest that additional work is required. This paper outlines the progression of work done by the heat transfer community over the last several decades as both the measurements and the predictions have improved to current levels. To frame the problem properly, the paper reviews the influence of turbine aerodynamics on heat transfer predictions. This includes a discussion of time-resolved surface-pressure measurements with predictions and the data involved in forcing function measurements. The ability of existing two-dimensional and three-dimensional Navier-Stokes codes to predict the proper trends of the time-averaged and unsteady pressure field for full-stage rotating turbines is demonstrated. Most of the codes do a reasonably good job of predicting the surface-pressure data at vane and blade midspan, but not as well near the hub or the tip region for the blade. In addition, the ability of the codes to predict surface-pressure distribution is significantly better than the corresponding heat transfer distributions. Heat transfer codes are validated against measurements of one type or another. Sometimes the measurements are performed using full rotating rigs, and other times a much simpler geometry is used. In either case, it is important to review the measurement techniques currently used. Heat transfer predictions for engine turbines are very difficult because the boundary conditions are not well known. The conditions at the exit of the combustor are generally not well known and a section of this paper discusses that problem. The majority of the discussion is devoted to external heat transfer with and without cooling, turbulence effects, and internal cooling. As the design community increases the thrust-to-weight ratio and the turbine inlet temperature, there remain many turbine-related heat transfer issues. Included are film cooling modeling, definition of combustor exit conditions, understanding of blade tip distress, definition of hot streak migration, component fatigue, loss mechanisms in the low turbine, and many others. Several suggestions are given herein for research and development areas for which there is potentially high payoff to the industry with relatively small risk.

Theoretical and Measured Electron-Density Distributions at High Altitudes

A viscous shock-layer analysis has been developed and applied to the calculation of nonequilibrium-flow species distributions in the plasma layer of a blunt-nosed vehicle at high altitudes. The theoretical electron-density results obtained are in good agreement with those measured in flight for a hemisphere-9 cone entry vehicle. The flight measurements were obtained using electrostatic probes that protruded well into the shock layer. In addition, the theoretically obtained heavy-particle translational temperatures appear to agree fairly well with the electron temperatures that were measured in the flight experiments using voltage-swept thin-wire electrostatic probes. The influence of the reaction-rate coefficients on the calculated electron densities has been assessed and shown to be within the uncertainty in the flight data. The theoretical results demonstrate the importance of including in the chemical model the positive ions N2 + , O2 + , N + , and O + , in addition to NO + , for the high altitudes and velocities considered here.

Vane-blade interaction in a transonic turbine. I - Aerodynamics

Part I of this article presents results of a computational investigation of the effects of blade row interaction on the aerodynamics of a transonic turbine stage. The predictions are obtained using a two-dimensional unsteady Navier-Stokes code based on an explicit Runge-Kutta algorithm and an overlapping O-H grid system. This code simulates the flow in time-accurate fashion using nonreflective stage inflow and outflow boundary conditions and phase-lagging procedures for modeling arbitrary airfoil counts in the vane and blade rows. The O-H grid provides high spatial resolution of the high gradient regions near the airfoil surfaces and allows for arbitrary placement of stage inflow and outflow boundaries. Unsteady and time-averaged airfoil surface pressure predictions are compared with those from an older version of the code based on the explicit hopscotch algorithm and an O-grid system, and experimental data obtained in a short-duration shock tunnel facility.

Time-Averaged Heat Flux for a Recessed Tip, Lip, and Platform of a Transonic Turbine Blade

The results of an experimental research program determining the blade platform heat-flux level and the influence of blade tip recess on the tip region heat transfer for a full-scale rotating turbine stage at transonic vane exit conditions are described. The turbine used for these measurements was the Allison VBI stage operating in the closed vane position (vane exit Mach number1.1). The stage was operated at the design flow function, total to static pressure ratio, and corrected speed. Measurements were obtained at several locations on the platform and in the blade tip region. The tip region consists of the bottom of the recess, the lip region (on both the pressure and suction surface sides of the recess), and the 90 percent span location on the blade suction surface. Measurements were obtained for three vane/blade spacings: 20, 40, and 60 percent of vane axial chord and for a single value of the tip gap (the distance between the top of the lip and the stationary shroud) equal to 0.0012 m (0.046 in) or 2.27 percent of blade height.

Operation of gas turbine engines in volcanic ash clouds

Results are reported for a technology program designed to determine the behavior of gas turbine engines when operating in particle-laden clouds. There are several ways that such clouds may be created, i.e., explosive volcanic eruption, sand storm, military conflict, etc. The response of several different engines, among them the Pratt & Whitney JT3D turbofan, the Pratt & Whitney J57 turbojet, a Pratt & Whitney engine of the JT9 vintage, and an engine of the General Electric CF6 vintage has been determined. The particular damage mode that will be dominant when an engine experiences a dust cloud depends upon the particular engine (the turbine inlet temperature at which the engine is operating when it encounters the dust cloud), the concentration of foreign material in the cloud, and the constituents of the foreign material (the respective melting temperature of the various constituents). Further, the rate at which engine damage will occur depends upon all of the factors given above, and the damage is cumulative with continued exposure. An important part of the Calspan effort has been to identify environmental warning signs and to determine which of the engine parameters available for monitoring by the flight crew can provide an early indication of impending difficulty. On the basis of current knowledge, if one knows the location of a particle-laden cloud, then that region should be avoided. However, if the cloud location is unknown, which is generally the case, then it is important to know how to recognize when an encounter has occurred and to understand how to operate safely, which is another part of the Calspan effort.

Experimental Results From Controlled Blade Tip/Shroud Rubs at Engine Speed

Experimental results obtained for an Inconel® compressor blade rubbing a steel casing at engine speed are described. Load cell, strain gauge, and accelerometer measurements are discussed and then applied to analyze the metal-on-metal interaction resulting from sudden incursions of varying severity, defined by incursion depths ranging from 13 μm to 762 jam (0.0005 in. to 0.030 in.). The results presented describe the transient dynamics of rotor and casing vibro-impact response at engine operational speed similar to those experienced in flight. Force components at the blade tip in axial and circumferential directions for a rub of moderate incursion depth (140 μm) are compared to those for a severe rub (406 μm). Similar general trends of variation during the metal-to-metal contact are observed. However, in the nearly threefold higher incursion the maximum incurred circumferential load increases significantly, while the maximum incurred axial load increases much less, demonstrating the non-linear nature of the rub phenomena. Concurrently, the stress magnification on the rubbing blade at root mid-chord, at tip leading edge, and at tip trailing edge is discussed. The results point to the possibility of failure occurring first at the airfoil trailing edge. Such a failure was in fact observed in the most severe rub obtained to date in the laboratory, consistent with field observations. Computational models to analyze the non-linear dynamic response of a rotating beam with periodic pulse loading at the free-end are currently under development and are noted.

Interpretation of Gas Turbine Response Due to Dust Ingestion

A measurement program currently underway at Arvin/Calspan Advanced Technology Center has been used in the evaluation of observed engine behavior during dust ingestion. The Pratt and Whitney TF33 turbofan and J57 turbojet were used in the investigation. Solid particle ingestion was found to erode the compressor blades and result in substantial performance deterioration. The engines were found to have increased susceptibility to surge at low power settings. The roles that anti-ice and intercompressor bleed airplay in surge avoidance are discussed. A discussion of the fuel controller behavior in a deteriorated engine and its effect during steady-state engine operation is also presented. Experimental data obtained during testing were compared to a predictive capability developed to describe deteriorated engine response. The effects of tip clearance, blade profile, and secondary flows were taken into account. The results show good agreement with experimentally observed engine behavior.

Development of an experimental capability to produce controlled blade tip/shroud rubs at engine speed

An experimental capability using an in-ground spin-pit facility specifically designed to investigate aeromechanic phenomena for gas turbine engine hardware rotating at engine speed is demonstrated herein to obtain specific information related to prediction and modeling of blade-casing interactions. Experiments are designed to allow insertion of a segment of engine casing into the path of single-bladed or multiple-bladed disks. In the current facility configuration, a 90 deg sector of a representative engine casing is forced to rub the tip of a single-bladed compressor disk for a selected number of rubs with predetermined blade incursion into the casing at rotational speeds in the vicinity of 20,000 rpm.

Measurement and Analyses of Heat Flux Data in a Turbine Stage: Part II—Discussion of Results and Comparison With Predictions

Mesure et analyse des donnees du flux thermique dans un etage de turbine. Discussion des resultats et comparaison avec les valeurs calculees

Measurement of Heat Flux and Pressure in a Turbine Stage

This paper describes heat-flux and pressure mesurements that were obtained using state-of-the-art, shock-tube technology and well established transient-test techniques (thin-film gages and fast-response pressure transducers). These measurements were performed for a full turbine stage of the AiResearch TFE 731-2 engine. 27 refs.