No description available.

Thesis (Ph. D. in Aeronautics and Astronautics)--Naval Postgraduate School, June 2002. / Dissertation supervisor: Raymond P. Shreeve. Includes bibliographical references (p. 157-161). Also available online.

Nearly periodic structures exhibit localized modes of vibration depending upon the ex· tent of the coupling and the disorder between their periodic components. These localized modes may cause excessive vibration amplitudes resulting in significantly higher stresses than the ones the structures have been designed to withstand. The Rayleigh-Ritz method, as employed earlier, for dynamic analysis for two-span beams could predict only the lower modes of vibration at prohibitive computational expense. A semi-analytical method which predicts even the higher modes at significantly reduced cost has been applied to the free vibration analysis of two-span beams. This method has been applied to the forced vibration analysis of a couple of mistuned turbine blades modeled as a two- span beam rotating in a turbulent flow field. It is shown that the maximum vibration amplitude of the mistuned structure may increase by several hundred percent as compared to that of the perfectly ordered structure. / Master of Science

No description available.

Tip leakage losses were studied in a linear turbine cascade with a tip clearance gap equal to 2.1 percent of blade height. The blades of the cascade have a turning angle of 109.4 degrees, an aspect ratio of 1.0, and an axial chord length of 235.2 mm. The cascade was located at the exit of a low speed wind tunnel; the blade exit Reynolds number based upon blade axial chord was 4.5x10⁵. The flow was measured at a plane 0.96 axial chords downstream from the blade leading edge. Barlier studies performed at the tip gap exit and at a downstream plane 1.4 axial chords from the blade leading edge were utilized with the present study to understand loss development better. The effect of tip leakage and the corresponding loss production mechanisms involved as the flow mixes out were analyzed. As part of the objective of the study, a computerized data acquisition system was developed which acquires pressure data and controls movement of a five hole pressure probe. The flow properties at the measurement plane were numerically integrated. To estimate the maximum potential loss of the cascade, the flow was mixed-out through a momentum analysis. The loss at the measurement plane due to tip leakage was found to be equal to the sum of the total pressure loss within the tip gap and the dissipated tip gap secondary kinetic energy. As the flow proceeded downstream, losses were attributed to dissipation of secondary kinetic energy, trailing edge wake mixing, endwall losses, and primary flow mixing. / Master of Science

Thesis (MEng.)-- University of Stellenbosch, 1992. 140 leaves on single pages, preliminary pages i-xi and numbered pages 1-113. Includes bibliography. Digitized at 600 dpi grayscale to pdf format (OCR),using an Bizhub 250 Konica Minolta Scanner and at 300 dpi grayscale to pdf format (OCR), using a Hp Scanjet 8250 Scanner. / Thesis (MEng (Mechanical and Mechatronic Engineering))--University of Stellenbosch, 1992 / ENGLISH ABSTRACT: The large variety of turbo-machinery blade root geometries in use in industry prompted the question if a optimum geometry could be found. An optimum blade root was defined as a root with a practical geometry which, when loaded, returns the minimum fillet stress concentration factor. A literature survey on the subject provided guidelines but very little real data to work from. An initial optimization was carried out using a formula developed by Heywood to determine loaded projection fillet stresses. The method was found to produce unsatisfactory results, prompting a photoelastic investigation. This experimental optimization was conducted in two stages. A single tang defined load stage and a single tang in-rotor stage which modeled the practical situation. The defined load stage was undertaken in three phases. The first phase was a preliminary investigation, the second phase was a parameter optimization and the third phase was a geometric optimization based on a material utilization optimization. This material optimization approach produced good results. From these experiments a practical optimum geometry was defined. A mathematical model which predicts the fillet stress concentration factor for a given root geometry is presented. The effect of expanding the single tang optimum to a three tang root was examined. / AFRIKAANSE OPSOMMING: Die groot verskeidenheid lemwortelgeometrieë wat in turbomasjiene gebruik word het die vraag na 'n optimum geometrie laat ontstaan. Vir hierdie ondersoek is 'n optimum geometrie gedefineer as 'n praktiese geometrie wat, as dit belas word, die mimimum vloeistukspanningskonsentrasiefaktor laat ontstaan. 'n Literatuur studie het riglyne aan die navorsing gegee maar het wynig spesifieke en bruikbare data opgelewer. Die eerste optimering is met die Heywood formule, wat vloeistukspannings in belaste projeksies bepaal, aangepak. Die metode het nie bevredigende resultate opgelewer nie. 'n Fotoelastiese ondersoek het die basis vir verdere optimeering gevorm. Hierdie eksperimentele optimering is in twee stappe onderneem. 'n Enkelhaak gedefineerde lasgedeelte en 'n enkelhaak in-rotor gedeelte het die praktiese situasie gemodeleer. Die gedefineerde lasgedeelte is in drie fases opgedeel. Die eerste fase was n voorlopige ondersoek. Die tweede fase was 'n parameter optimering. 'n Geometrie optimering gebasseer op 'n materiaal benuttings minimering het die derde fase uitgemaak. Die materiaal optimerings benadering het goeie resultate opgelewer. Vanuit hierdie eksperimente is 'n optimum praktiese geometrie bepaal. 'n Wiskundige model is ontwikkel, wat die vloeistukspanningskonsentrasiefaktor vir 'n gegewe wortelgeometrie voorspel. Die resultaat van 'n geometriese uitbreiding van die enkelhaaklemwortel na 'n driehaaklemwortel op die spanningsverdeling is ondersoek.

The influence of tip clearance and angle of attack upon the mid-span unsteady pressure response of a vibrating flat plate was investigated experimentally. Unsteady pressure measurements were taken for a variety of incidence angles, vibration frequencies and tip clearances over a Mach number range of 0.2 to 0.6. It was found that changes in tip clearance had an effect on measured pressure fluctuations at higher angles of attack and larger Mach numbers. It was also observed that the amplitude of the unsteady pressure increased as the incidence angle was increased. The plate was mechanically induced to oscillate in translation, simulating the flISt bending mode. Averaged Fast Fourier Transforms were used to determine pressure oscillation amplitudes and phase lags with respect to the plate motion. / Master of Science

No description available.

Highly-loaded turbomachine blades can stall under off-design conditions. In this regime, the flow can separate close to the leading edge of the blade in a periodic manner that can lead to blade vibrations, commonly referred to as stall flutter. Prior experimental studies on stall flutter have been at large Re (Re ~ 106). In the present work, motivated by applications in Unmanned Air Vehicles (UAV) and Micro Air Vehicles (MAV), we study experimentally the forces and flow fields around an oscillating blade at low Re (Re ~ 3 x 104). At these low Re, the flow even over the stationary blade can be quite different. We experimentally study the propensity of an isolated symmetric and cambered blade (with chord c) to undergo self-excited oscillations at high angles of attack and at low Reynolds numbers (Re ~ 30, 000). We force the blade, placed at large mean angle of attack, to undergo small amplitude pitch oscillations and measure the unsteady loads on the blade. From the measured loads, the direction and magnitude of energy transfer to/from the blade is calculated. Systematic measurements have been made for varying mean blade incidence angles and for different excitation amplitudes and frequencies (f). These measurements indicate that post stall there is a possibility of excitation of the blade over a range of Strouhal Numbers (St = fc/U) with the magnitude of the exciting energy varying with amplitude, frequency and mean incidence angles. In particular, the curves for the magnitude of the exciting energy against Strouhal number (St) are found to shift to higher St values as the mean angle of attack is increased. We perform the same set of experiments on two different blade shapes, namely NACA 0012 and a compressor blade profile, SC10. Both blade profiles show qualitatively similar phenomena. The flow around both the stationary and oscillating blades is studied through Particle Image Velocimetry (PIV). PIV measurements on the stationary blade show the gradual shift of the flow separation point towards the leading edge with increasing angle of attack, which occurs at these low Re. From PIV measurements on an oscillating blade near stall, we present the flow field around the blade at different phases of the blade oscillation. These show that the boundary layer separates from the leading edge forming a shear layer, which flaps with respect to the blade. As the Strouhal number is varied, the phase between the flapping shear layer and the blade appears to change. This is likely to be the reason for the observed change in the sign of the energy transfer between the flow and the blade that is responsible for stall flutter.