An experimental study was performed to investigate the mechanics of fluid flow in a 30<SUP>o</SUP> annular diffuser, and to study the periodicity of flow oscillations in this region. Surface oil-flow patterns and smoke-flow visualisation experiments were performed with a centrebody concentric and eccentric in the diffuser. Above a threshold offset, the measurements revealed two contra-rotating periodic vortices in antiphase with each other, symmetrically disposed about the plane of minimum clearance in the annulus, and originating in the 30<SUP>o</SUP> diffuser. Steady pressure measurements indicated that the steady fluid forces acting on the centre body are decentralising, and any small perturbation will result in the centrebody being pushed towards the wall of the diffuser. Unsteady pressure and force measurements showed that for a mean eccentric centrebody position, there was a predominantly tangential unsteady vortex force present with the centre body both fixed and vibrating. This vortex force scaled linearly with flow velocity, indicating a Strouhal-type mechanism. The magnitude of the vortex force was independent of both amplitude and frequency of vibration of the centrebody, indicating a forced vibration effect. The exception was when the centre body frequency approached the vortex shedding frequency, in which case lock-on occurred. For the geometry considered, lock-on does <i>not</i> significantly increase the unsteady forces acting on the centrebody. During lock-on it was found that the vortices could not only be influenced by centrebody motion, but could be completely suppressed by closely controlling the amplitude and frequency of the centrebody. The effect that shaking the centrebody has on the different flow regimes in which the annular diffuser operates is explained. The vortices could also be eliminted by a) attaching a small helical fence to the surface of the centrebody, and b) by inserting a perforated liner within, and downstream of, the diffuser section. A small perturbation theoretical analysis of the unsteady flow in the diffuser has been developed. The flow was computed numerically, and the predicted self-induced forces examined. The analysis predicted mainly negative damping for the configurations examined. The predicted magnitude of the unsteady forces agreed with experimental results. Finally, the flow was also predicted analytically, and a good level of agreement with the numerical study was found.

A subsonic wind tunnel facility was built and tested as part of a base line test investigating flow within a diffuser-collector. Facility controls allowed the quarter scale model to match both Reynolds number and Mach number. Mass averaged conditions at the diffuser inlet during testing were determined as 1.939 ? 106 for Reynolds number based upon diffuser inlet hydraulic diameter, and 0.418 for Mach number. A flow conditioning section prior to test section contained several interchangeable sections. Flow conditioning components were used to create flow characteristic of that leaving the last stage of a land-based gas turbine. The diffuser-collector subsystem was evaluated through the use of wall static pressure measurements, a variety of probe traverse measurements, and Stereo-PIV. Flow within the collector and diffuser were determined to be heavily dependent upon the collector geometry. PIV measurements showed the development of two large counter rotating vortices within the collector. Each symmetric vortex grew and shifted according to the collector geometry while creating complex regions of flow. Pressure recovery within the diffuser was in range of 0.47 to 0.78, and would drop to 0.52 at the collector exit. The drop in pressure recovery was presumed to be a combination of inefficient diffusion in the collector and losses due to the vortices. The baseline test was found to be successful in terms of facility design, and determining the critical flow phenomena. Further testing and experimentation are necessary to evaluate specific details of the collector geometry's effect upon the pressure recovery and flow development. / Master of Science

Short annular diffusers were essential components for turbomachines that have been used to expand the air entering the compressor, as interstage ducts between gas generators and power turbines, and on the exhaust gases exiting the turbine. The industrial community was interested and invested in improving diffuser design that was challenging owing to the unfavourable fluid flow effects. Efficient design of fluid flow devices was possible through the complementary use of experimental testing and computational fluid dynamics (CFD). A numerical shape optimization study was undertaken to determine preferential annular diffuser configurations. Experimental data were compared against CFD that simulated the steady-state Reynolds-averaged Navier-Stokes equations with two-equation turbulence models. This investigation reached equivalent conclusions with respect to the influences associated with diffuser geometry and swirl. Vorticity effects caused by square tabs, that were not as well understood, were investigated. The tabs were effective in reducing the central toroidal recirculation zone created by a swirling flow, but at a static pressure penalty for the area ratio, AR<2.73, diffusers tested. Results identified several shortcomings in the CFD that typically over-estimated pressure recovery and outlet velocity uniformity; however, properly qualitatively predicted wall pressure distributions and outlet velocity profiles. The use of CFD on modest grids, with preference given to the realizable k-epsilon turbulence model, for annular diffusers that have length to inlet height ratio of 12 and at least AR=2.73 with up to 20-degrees inlet swirl was encouraged as a design tool. / Thesis (Ph.D, Mechanical and Materials Engineering) -- Queen's University, 2014-05-29 09:03:16.591

The combustor-diffuser system remains one of the most studied sections of the turbomachine. Most of these investigations are due to the fact that quite a bit of flow diffusion is required in this section as the high speed flow exits the compressor and must be slowed down to enter the combustor. Like any diffusion process there is the chance for the development of an unfavorable adverse pressure gradient that can lead to flow separation; a cause of drastic losses within a turbine. There are two diffusion processes in the combustor-diffuser system: The flow first exits the compressor into a pre-diffuser, or compressor discharge diffuser. This diffuser is responsible for a majority of the pressure recovery. The flow then exits the pre-diffuser by a sudden expansion into the dump diffuser. The dump diffuser comprises the majority of the losses, but is necessary to reduce the fluid velocity within acceptable limits for combustion. The topic of active flow control is gaining interest in the industry because such a technique may be able to alleviate some of the requirements of the dump diffuser. If a wider angle pre-diffuser with separation control were used the fluid velocity would be slowed more within that region without significant losses. Experiments were performed on two annular diffusers to characterize the flow separation to create a foundation for future active flow control techniques. Both diffusers had the same fully developed inlet flow condition, however, the expansion of the two diffusers differed such that one diffuser replicated a typical compressor discharge diffuser found in a real machine while the other would create a naturally separated flow along the outer wall. Both diffusers were tested at two Reynolds numbers, 5x104 and 1x105, with and without a vertical wall downstream of the exit to replicate the dump diffuser that re-directs the flow from the pre-diffuser outlet to the combustor. Static pressure measurements were obtained along the OD and ID wall of the diffusers to determine the recovered pressure throughout the diffuser. In addition to these measurements, tufts were used to visualize the flow. A turbulent CFD model was also created to compare against experimental results. In the end, the results were validated against empirical data as well as the CFD model. It was shown that the location of the vertical wall was directly related to the amount of separation as well as the separation characteristics. These findings support previous work and help guide future work for active flow control in a separated annular diffuser both computationally and experimentally.