Computational fluid dynamics or CFD isan importanttool thatis used at various stages in the design of highly complex turbomachinery such as compressorand turbine stages that are used in land and air based power generation units. The ability of CFD to predict the performance characteristics of a specific blade design is challenged by the need to use various turbulence models to simulate turbulent flows as well as transition models to simulate laminar to turbulent transition that can be observed in various turbomachinery designs. Moreover, CFD is based on numerically solving highly complex differential equations, which through the use of a grid to discretize the geometry introduces numerical errors. Allthese factors combine to challenge CFD’s role as a predictor of blade performance. It has been generallyfound that CFD in its current state of the art is best used to compare between various design points and not as a pure predictor of performances. In this study the capability of CFD, and turbulence modeling, in turbomachinery based geometry is assessed.Three different blade designs are tested, that include an advanced two-stage turbine blade design, a three stage 2D or cylindrical design and finally a three stage bowed stator and rotor design. Allcases were experimentally tested at the Texas A&Muniversity Turbomachinery Performance and Flow Research Laboratory (TPFL).In all cases CFD provided good insights into fundamental turbomachinery flow physics, showing the expected improvement from using 2D cylindrical blades to 3D bowed blade designs in abating the secondary flow effects which are dominant loss generators.However, comparing experimentally measured performance results to numerically predicted shows a clear deficiency, where the CFD overpredicts performance when compared to experimentallyobtained data, largely underestimating the various loss mechanisms. In a relative sense, CFD as a tool allows the user to calculate the impact a new feature or change can have on a baseline design. CFD will also provide insight into what are the dominant physics that explain why a change can provide an increase or decrease in performance. Additionally,as part of this study, one of the main factors that affect the performance of modern turbomachinery is transition from laminar to turbulent flow.Transition is an influential phenomena especially in high pressure turbines, and is sensitive to factors such asupstream incidentwake frequency and turbulence intensity.A model experimentally developed, is implemented into a CFD solver and compared to various test results showing greater capability in modeling the effects of reduced frequency on the transition point and transitional flow physics. This model is compared to industry standard models showing favorable prediction performance due to its abilityto account for upstream wake effects which most current model are unable to account for.

Engine development has been in the direction of increased turbine inlet temperatures to improve efficiency and power output. Secondary flows develop as a result of a near-wall pressure gradient in the stagnating flow approaching the inlet nozzle guide vane as well as a strong cross-passage gradient within the passage. These flow structures enhance heat transfer and convect hot core flow gases onto component surfaces. In modern engines it has become critical to cool component surfaces to extend part life. Bypass leakage flow emerging from the slot between the combustor and turbine endwalls can be utilized for cooling purposes if properly designed. This study examines a three-dimensional slot geometry, scalloped to manipulated leakage flow distribution. Statistical techniques are used to decouple the effects of four geometric parameters and quantify the relative influence of each on endwall cooling levels and near-wall total pressure losses. The slot geometry is also optimized for robustness across a range of inlet conditions. Average upstream distance to the slot is shown to dominate overall cooling levels with nominal slot width gaining influence at higher leakage flow rates. Scalloping amplitude is most influential to near-wall total pressure loss as formation of the horseshoe vortex and cross flow within the passage are affected. Scalloping phase alters local cooling levels as leakage injection is shifted laterally across the endwall. / Ph. D.

The flow and stability of co-extruded layers of different polymers in a forced assembly process is studied computationally to determine the extent of the stable process window and the types of instabilities that occur. Recent advances in layer-multiplying co-extrusion of incompatible polymers have made possible the fabrication of multilayered nanostructures with improved barrier, thermal and mechanical behavior. However, existing layering techniques are very sensitive to mismatches in viscosity and elasticity of the co-extruded polymers which often give rise to layer non-uniformity and flow instabilities, such as encapsulation. Simulations of the flows inside the feedblock and the successive multiplier dies of the multi-layering system are used to track the interface and predict instabilities and degrees of encapsulation as a function of process parameters, primarily the flow rates and rheology of the polymers. Encapsulation is found to be negligible in practice in the feedblock even for large viscosity contrasts and differences in elasticity between the two co-extruded polymers. Encapsulation or pinch-off of interfaces is more severe in the multiplier dies when there the rheologies of the polymers differ. A secondary flow due to the second normal stress differences for non-Newtonian fluids is primarily responsible for the encapsulation. A new multiplier design is proposed and simulated. The pressure drop in the proposed design is half that of the current design, which is useful for extruding highly elastic materials. Further, the degree of encapsulation is also reduced. The results of the simulations are validated with experimental measurements of pressure drop and flow visualization provided by research collaborators. / text

The possibility of misalignments at interfaces would be increased due to individual parts' assembly and external factors during its operation. In actual engine representative conditions, the upstream misalignments have effects on turbines performance through the nozzle guide vane passages. The current experimental aerodynamic investigation over the nozzle guide vane passage was concentrated on the backward-facing step of upstream misalignments. The tests were performed using two types of vane endwall platforms in a 2D linear cascade: flat endwall and axisymmetric converging endwall. The test conditions were a Mach number of 0.85, Re_ex 1.5*10^6 based on exit condition and axial chord, and a high freestream turbulence intensity (16%), at the Virginia tech transonic cascade wind tunnel. The experimental results from the surface flow visualization and the five-hole probe measurements at the vane-passage exit were compared with the two cases with and without the backward-facing step for both types of endwall platforms. As a main source of secondary flow, a horseshoe vortex at stagnation region of the leading edge of the vane directly influences other secondary flows. The intensity of the vortex is associated with boundary layer thickness of inlet flow. In this regard, the upstream backward-facing step as a misalignment induces the separation and attachment of the inlet flow sequentially, and these cause the boundary layer of the inlet flow to reform and become thinner locally. The upstream-step positively affects loss reduction in aerodynamics due to the thinner inlet boundary layer, which attenuates a horseshoe vortex ahead of the vane cascade despite the development of the additional vortices. And converging endwall results in an increase of the effect of the upstream misalignment in aerodynamics, since the inlet boundary layer becomes thinner near the vane's leading edge due to local flow acceleration caused by steep contraction of the converging endwall. These results show good correlation with many previous studies presented herein. / Master of Science

The effect of rotation on flow in a gas turbine blade cooling duct model is investigated experimentally. The present work consists of velocity measurements at different locations in a test section with a 180° bend with ribs on one wall. Three geometric rib parameters are considered; rib-rib spacing, rib orientation angle and rib blockage ratio. PIV is used for flow visualization and analysis. Along with the clean duct measurements, ribs with blockage ratios, b/h, of 0.25 and 0.125 were considered. The b/h = 0.25 cases have been comprehensively analyzed while the b/h = 0.125 cases have been studied at the post-bend region of the duct only. Reynolds number considered is in the range of approximately 5000-40,000 and the rotation speed is varied for a rotation number from 0 to 7. It is observed that rotation has a significant effect on secondary flows within the rotating duct. For blockage ratio, b/h = 0.25, at a constant Re and Ro, the RMS of fluctuations of velocity do not show large spatial variations with ribs or rib orientations. At higher Re, the value decreases in comparison to the low Re cases. The kinetic energy of fluctuations increases due to the presence of ribs, indicating better heat transfer for the ribbed duct, but do not show large variations with rib orientation angle. The fluctuations and kinetic energy show maximum values at the post-bend regions of the duct. The velocity fields and PDFs show a possible cause for e°cient heat transfer for the 45° rib arrangement as compared with the 90° rib cases. At high Ro, the absolute value of circulation has a large increase at the post-bend and thereafter there is a gradual decay at the exit for all cases. The ribs with blockage ratio of b/h = 0.125 showed no marked changes in circulation with changes in rib orientation angle, thus implying that the rib blockage plays a role in the generation of secondary flows, particularly in conjunction with rotation.

Linear sandbanks are located globally in areas where there are strong currents and an abundance of sand. In recent years, these sandbanks have become a strategic interest as a potential source of marine aggregates (sand and gravel) and mineral deposits. They also commonly reach the sea surface and thus pose a threat to navigation. Headland-associated linear sandbanks are a specific type of sandbank, which are located in the lee of coastal topographic features such as headlands and islands. Interaction between tidal currents and topographic features generate complex three-dimensional circulation patterns that significantly influence the distribution of sediments in the vicinity of the feature. Field and numerical model investigations of the three-dimensional flow structure have been undertaken on the Levillain Shoal, a headland-associated linear sandbank present in the lee of Cape Levillain (Shark Bay, Western Australia). The field data indicated the presence of secondary flows near the tip of the Cape and around the bank, which were reproduced in the numerical simulations. Sediment transport paths near the Cape and the bank indicate that the sandbank is part of a sand circulation cell where the sand is circulating around the bank with exchanges between the sandbank and the headland. A morphological model (MTM) has been developed to understand processes responsible for the formation of the headland associated linear sandbanks. With an “idealized” Gaussian shaped headland, the formation of two symmetrical sandbanks on each side of the headland is observed. It is shown that sandbanks are formed in regions where there is a net accumulation of sand over a tidal cycle, due to the acceleration/⁄deceleration effects of the flow in the presence of the headland. Initially, sandbanks develop in a circular shape and grow vertically. As the sandbanks interact with the tidal flow, they evolve into elongated linear deposits (as observed in nature). The sandbank growth is dependent on the tidal regime, secondary flow, sand availability, and sediment grain size

This paper presents details of an experimental and computational investigation on the effect of misalignment between the combustor exit and nozzle guide vane endwall on the heat transfer distribution across an axisymmetric converging endwall. The axisymmetric converging endwall investigated was representative of that found on the shroud side of a first stage turbine nozzle section. The experiment was conducted at a nominal exit M of 0.85 and exit Re 1.5 x 10⁶ with an inlet turbulence intensity of 16%. The experiment was conducted in a blowdown transonic linear cascade wind tunnel. Two different inlet configurations were investigated. The first configuration, Case I, was representative of a combustor exit aligned to the nozzle platform, with a gap located at the interface of the tow components. The second configuration, Case II, the endwall platform was offset in the span-wise direction to create a backward facing step at the inlet. This step is representative of a misalignment between the combustor exit and the NGV platform. An infrared camera was used to capture the temperature history on the endwall, from which the endwall heat transfer distribution was determined. A numerical study was also conducted by solving RANS equations using ANSYS Fluent v.15. The numerical results provided insight into the passage flow field which explained the observed heat transfer characteristics. Case I showed the typical characteristics of transonic vane cascade flow, such as the separation line, saddle point, and horseshoe vortices. The presence of a gap at the combustor-nozzle interface facilitated the formation of a separated flow which propagated through the passage. This flow feature caused the passage vortex reattach to the SS vane at 0.44 x/C. The addition of the platform misalignment in Case II caused the flow reattachment region to occur near the vane LE plane. The separated flow which formed at the inlet step, merged with the recirculation region on the endwall platform, forming two counter-rotating auxiliary vortices. These vortices significantly delayed migration of the passage vortex, causing it to reattach on the SS vane at 0.85 x/C. These two flow features also had a significant effect on the endwall heat transfer characteristics. The heat transfer levels on the endwall platform, from -0.50 to +0.50 Cx relative to the vane LE, had an average increase of ~40%. However, downstream of the vane mid-passage, the heat transfer levels showed no appreciable heat transfer augmentation due to flow acceleration through the passage throat. / Master of Science

The flow velocity in a river is three-dimensional (3D), turbulent, and varies in time and space. Capturing this variability in field measurements to support studies of river processes has proven particularly challenging. While originally developed to measure discharge, boat-mounted acoustic Doppler current profilers (ADCP) are increasingly used in field studies to quantify flow features including mean velocity, boundary shear stress, and sediment motion. Two survey procedures are typically employed with an ADCP. Moving-vessel (MV) measurements provide spatially-rich velocity data while temporally-rich data are obtained with fixed-vessel (FV) procedures. Given the relative ease of MV measurements, recent work has focused on developing MV procedures that produce comparable results to FV measurements. At the present, results of this work are inconclusive. Additionally, there is a lack of reported data and procedures for FV measurements. This work seeks to develop techniques to present 3D velocity data obtained in natural rivers in a unified framework. This framework is based on a stream-fitted coordinate system defined by the flow direction at a cross section and allows for 3D velocity to be decomposed into streamwise, spanwise, and vertical components. Procedures are developed to assure that the velocity profiles measured at fixed locations are (1) not negatively impacted by the inevitable motion of the ADCP, (2) statistically stationary, and (3) of sufficient record length to determine the mean velocity. The coordinate system allows time-averaged velocity from FV procedures to be compared with spatially-averaged velocity from MV vessels. Significant differences are found between the two survey procedures, particularly for secondary velocity components. Ultimately, integrating results of the two survey procedures leads to an improved representation of the mean flow field. The techniques are applied to data obtained on a study reach on the lower Roanoke River, located in eastern North Carolina. / Ph. D.

This study describes a detailed investigation on the effects that upstream step misalignment and upstream purge film cooling have on the endwall heat transfer for nozzle guide vanes in a land based power generation gas turbine at transonic conditions. Endwall Nusselt Number and adiabatic film cooling effectiveness distributions were experimentally calculated and compared with qualitative data gathered via oil paint flow visualization which also depicts endwall flow physics. Tests were conducted in a transonic linear cascade blowdown facility. Data were gathered at an exit Mach number of 0.85 with a freestream turbulence intensity of 16% at a Re = 1.5 x 106 based on axial chord. Varied upstream purge blowing ratios and a no blowing case were tested for 3 different upstream step geometries, one of which was the baseline (no step). The other two geometries are a backward step geometry and a forward step geometry, which comprised of a span-wise upstream step of +4.86% span and -4.86% span respectively. Experimentation shows that the addition of upstream purge film cooling increases the Nusselt Number at injection upwards of 50% but lowers it in the throat of the passage by approximately 20%. The addition of a backward facing step induces more turbulent mixing between the coolant and mainstream flows, thus reducing film effectiveness coverage and increasing Nusselt number by nearly 40% in the passage throat. In contrast, the presence of a forward step creates a more stable boundary layer for the coolant flow, thus aiding to help keep the film attached to the endwall at higher blowing ratios. Increasing the blowing ratio increases film cooling effectiveness and endwall coverage up to a certain point, beyond which, the high momentum of the coolant results in poor cooling performance due to jet liftoff. Near endwall streamlines without purge cooling generated by Li et al. [1] for the same geometries were compared to the experimental data. It was shown that even with the addition of upstream purge cooling, the near endwall streamlines as they moved downstream matched strikingly well with the experimental data. This discovery indicates that while the coolant flow will likely affect the flow streamlines three dimensionally, they are minimally effected by the coolant flow near the endwall as the flow moves downstream. / Master of Science

This dissertation presents a detailed experimental and numerical analysis of the aerothermal characteristics of the turbine extremity regions i.e. the blade tip and endwall regions. The heat transfer and secondary flow characteristics were analyzed for different engine relevant configurations and exit Mach/Reynolds number conditions. The experiments were conducted in a linear blowdown cascade at transonic high turbulence conditions of Mexit ~ 0.85, 0.60 and 1.0, with an inlet turbulence intensity of 16% and 12% for the vane and blade cascade respectively. Transient infrared (IR) thermography technique and surface pressure measurement were used to map out the surface heat transfer coefficient and aerodynamic characteristics. The experiments were complemented with computational modeling using the commercial RANS equation solver ANSYS Fluent. The CFD results provided further insight into the local flow characteristics in order to elucidate the flow physics which govern the measured heat transfer characteristics. The results reveal that the highest heat transfer exists in regions with local flow reattachment and new-boundary layer formation. Conversely, the lowest heat transfer occurs in regions with boundary layer thickening and separation/lift-off flow. However, boundary layer separation results in additional secondary flow vortices, such as the squealer cavity vortices and endwall auxiliary vortex system, which significantly increase the stage aerodynamic losses. Furthermore, these vortices result in a low film-cooling effectiveness as was observed on a squealer tip cavity with purge flow. Finally, the importance of transonic experiments in analyzing the turbine section heat transfer and flow characteristics was underlined by the significant shock-boundary layer interactions that occur at high exit Mach number conditions. / Ph. D.