Showing papers by "Terrence W. Simon published in 2003"

Characterizing the performance of the sr-30 turbojet engine

Abstract: The SR-30 is a small-scale, turbojet engine which sounds and smells like a real engine used to fly commercial aircraft. With an overall length of less than 2.0 feet and an average diameter of 6.5 inches, the SR-30 is equipped with an inlet nozzle, radial compressor, counter-flow combustion chamber, turbine, and exhaust nozzle. Although it can operate on various fuels, diesel fuel is used in the studies described here, and each component is instrumented with thermocouples and pressure gages to allow a complete thermodynamic evaluation. Screaming along at 80,000 rpm and sending out exhaust gas at 500 mph, the SR-30 engine is fun science for students! However, since the SR-30 was essentially designed for one-dimensional measurement and flow analysis, students quickly learn the limitations of these assumptions.

Overview of NASA Multi-dimensional Stirling Convertor Code Development and Validation Effort

Abstract: A NASA grant has been awarded to Cleveland State University (CSU) to develop a multi-dimensional (multi-D) Stirling computer code with the goals of improving loss predictions and identifying component areas for improvements. The University of Minnesota (UMN) and Gedeon Associates are teamed with CSU. Development of test rigs at UMN and CSU and validation of the code against test data are part of the effort. The one-dimensional (1-D) Stirling codes used for design and performance prediction do not rigorously model regions of the working space where abrupt changes in flow area occur (such as manifolds and other transitions between components). Certain hardware experiences have demonstrated large performance gains by varying manifolds and heat exchanger designs to improve flow distributions in the heat exchangers. 1-D codes were not able to predict these performance gains. An accurate multi-D code should improve understanding of the effects of area changes along the main flow axis, sensitivity of performance to slight changes in internal geometry, and, in general, the understanding of various internal thermodynamic losses. The commercial CFD-ACE code has been chosen for development of the multi-D code. This 2-D/3-D code has highly developed pre- and post-processors, and moving boundary capability. Preliminary attempts at validation of CFD-ACE models of MIT gas spring and "two space" test rigs were encouraging. Also, CSU's simulations of the UMN oscillating-flow fig compare well with flow visualization results from UMN. A complementary Department of Energy (DOE) Regenerator Research effort is aiding in development of regenerator matrix models that will be used in the multi-D Stirling code. This paper reports on the progress and challenges of this

A model of 90 degree turn oscillatory flow

Abstract: It seems that we need a model that will accommodate laminar flow in the stagnation region and turn into a turbulent model some distance away from this region. A single model with transition capabilities will be even better. A 2-D axisymmetric CFD (Computational Fluid Dynamics) model was developed to simulate the UMN test rig with a 90-degree flow turn to a radial flow in a channel between two discs. One case has been examined in this paper, S (Disc spacing) = 0.127 m and ReD =7600 for both unidirectional and oscillatory flows.

Unsteady fluid dynamics simulation of a stirling engine heater head

Abstract: This paper presents hot-wire anemometry data from an experiment that replicates important features of oscillatory flows in Stirling engines. These features include impinging and sink flows, large-scale separation zones, recirculation bubbles, unsteady shear layers and spatially varying transition and relaminarization zones. In addition to a characterization of oscillatory flow, this paper presents and compares results gathered in a unidirectional investigation to determine when, during the oscillation cycle, quasi-steady results can be applied to Stirling engine design. This paper is part of a program that focuses on characterization of fluid flow and heat transfer within Stirling engines. The unsteady velocity data presented herein serves to support the development of 3-D CFD models and verify the conditions under which 1-D systems models may be applied. The ultimate goal is that Stirling engine design codes can be used with a greater degree of confidence.

Micro- or Small- Gas Turbines

Abstract: Recently, increased attention has been paid to small gas turbine units. Microturbines with fuel cells and microturbines with heaters in a Combined Heat and Power (CHP) arrangement are showing great promise for supplying distributed electric and shaft power and process or space heat. Small jet engines are becoming increasingly popular for small piloted and non-piloted aircraft. In developing the microturbine for power generation, considerable attention has been paid to improving the recuperator, which had not seen such intensive research and development activity since the years of automobile gas turbine development. Also, the small size of the microturbine has led to a preference for radial turbomachinery, including radial inflow turbines. Smaller turbomachinery and other components have led to greater interest in ceramic components and ceramic coatings. Renewed interest in radial flow turbomachinery revealed a need to better understand radial-flow component fluid mechanics such as the effects of streamline curvature on boundary layer flows. Finally, small sizes have raised attention to low Reynolds number effects, such as transition and flow separation, particularly in the development of small axial-flow turbines for use in aircraft propulsion. In this report, we review recent developments documented in the literature for these areas of interest to small gas turbine engines and comment on contributions from our research lab.

Modeling Laminar-to-Turbulent Transition in a Low-Pressure Turbine Flow Which Is Unsteady Due to Passing Wakes: Part II — Transition Path

Abstract: The open literature on modeling transition to turbulence of boundary layer flows in low-pressure turbines is reviewed. Included are the separated flow transition onset models of Mayle and of Davis et al. and the attached flow transition onset models of Mayle, Abu-Ghannam and Shaw and Drela. Their results are applied against data previously taken at the University of Minnesota (UMN) to assess their performance for use on the suction surface of a turbine blade in the presence of passing wakes. The data show measurements of velocity, turbulence level, intermittency and spatial and temporal acceleration resolved in space and phase angle within the wake passing period. In a “quasi-static” comparison, the input values to the models are values taken from the experiment, resolved in phase angle within the wake-passing period. Predictions from the models (the flow states with regard to transition throughout the wake-passing period) are compared with instantaneous intermittency values taken from the experiment. The Mayle separated and attached flow onset models are shown to be successful for the case investigated when applied in that fashion. The Abu-Ghannam and Shaw and Drela transition onset models predict onset locations which are somewhat downstream of where the data indicate the transition onset to be. Unique characteristics regarding transition observed at different times in a wake passing cycle are discussed. Some reasons are given to explain the differences between experimental results and model predictions. Transition onset modeling is addressed in the present paper and transition path modeling is addressed in a companion paper (part II).Copyright © 2003 by ASME

The influence of unsteady acceleration and turbulence intensity on transition in low-pressure turbines

Abstract: Wakes generated by upstream airfoil rows cause an unsteady flow field at the downstream rows in a Low Pressure Turbine (LPT). The pressure gradients and the local turbulence intensity levels experienced by the airfoil surface boundary layers change as the wakes pass through the passage. Laminar-toturbulent transition onset, transition path characteristics and separation and reattachment locations (if the flow separates) also change during the wake passing cycle. In this paper, the effects on transition onset and transition path of unsteady acceleration (composed of spatial acceleration and temporal acceleration), as well as turbulence intensity, are described. Results come from an experiment in which the wakes and passage flow of an LPT are simulated. Spatial acceleration and temporal acceleration are influenced by both the geometry and the instantaneous flow field. Thus, both are dynamic, driven by the change in streamwise momentum associated with the passing wakes. The total acceleration is the sum of the spatial acceleration and the temporal acceleration. Since the boundary layer responds to the instantaneous total acceleration field, this is the term to monitor. However, it is instructive to study its two components separately. Another feature of the wakes which is important to transition is the turbulence field. The local turbulence intensity increases then drops as the wake passes overhead. Deceleration and high turbulence levels, as experienced by the boundary layer as the wake approaches, promote transition whereas acceleration and a low level of turbulence, as experienced as the wake departs, can delay transition. Also, it is well documented that immediately behind a strip of transitional flow (a turbulent strip or turbulent spot), the boundary layer is especially calm (the calmed region), a feature which adds to the dynamics of transition in the LPT. Nomenclature