Master of Science, The Ohio State University, 2020, Aero/Astro Engineering

This thesis covers the intensive research effort to elucidate the role of elevated temperature in deposition. Several experimental campaigns were conducted in this pursuit. The testing explored high temperature deposition with 0-10 micron Arizona Road Dust (ARD) with the intent of creating a yield strength model that included temperature effects and could be incorporated into the existing OSU deposition model. Experimental work was first conducted in the impulse kiln facility where small amounts of the test dust were placed on ceramic targets and rapidly exposed to temperatures between 1200K and 1500K. Trends in the packing factor confirmed the existence of two threshold values (1350K and 1425K) that could be linked to strength characteristics of the dust when exposed to high temperatures. Using the information obtained from the kiln experiments, HTDF testing was conducted between 1325K and 1525K. Exit temperatures were set at 25K intervals in this region with a constant jet velocity of 150 m/s. The capture efficiency data showed this trend with temperature and indicated a softening temperature and melting temperature of 1362K and 1512K respectively. With these critical values in hand, the Ohio State University Molten Model was created to modify yield strength with particle velocity and temperature. The model was tested using CFD and showed a good capability for capturing particle temperature effects in deposition from an impinging particle-laden jet. A subsequent test campaign was conducted to explore the effect of varying surface temperature on deposition. Hastelloy coupons with Thermal Barrier Coatings (TBCs) were subjected to a constant jet at 1600K jet and 200 m/s while being cooled via a backside impingement jet. Surface temperatures between 1455K and 1125K were impacted with 0-10 micron ARD while an IR camera monitored the surface. Coupons with higher coolant flowrates (lower surface temperature) saw significantly lower deposition rates than the higher surf (open full item for complete abstract)

Committee: Jeffrey Bons Dr. (Advisor); Randall Mathison Dr. (Committee Member) Subjects: Aerospace Engineering

Master of Science, The Ohio State University, 2012, Aero/Astro Engineering

An accelerated deposition test facility was used to study the relationship between film cooling, surface temperature, and particle temperature at impact on deposit formation. Tests were run at gas turbine representative inlet Mach numbers (0.1) and temperatures (1090°C). Deposits were created from lignite coal fly ash with median diameters of 1.3 and 8.8µm. Two CFM56-5B nozzle guide vane doublets, comprising three full passages and two half passages of flow, were utilized as the test articles. Tests were run with different levels of film cooling back flow margin and coolant temperature. Particle temperature upon impact with the vane surface was shown to be the leading factor in deposition. Since the particle must traverse the boundary layer of the cooled vane before impact, deposition is directly affected by the film and metal surface temperature as well. Film coolant jet strength showed only minor effect on deposit patterns on the leading edge. However, larger Stokes number (resulting in higher particle impact temperature) corresponded with increased deposit coverage area on the shower head region. Additionally, infrared measurements showed a strong correlation between regions of greater deposits and elevated surface temperature on the pressure surface. Thickness distribution measurements also highlighted the effect of film cooling by showing reduced deposition immediately downstream of cooling holes. A set of secondary tests were also conducted to briefly study the effect of Stokes number on leading edge deposition with no cooling, in order to support conclusions from the primary tests. It was found that larger Stokes number led to an increase in rate of deposition due to a greater number of particles being able to follow their inertial trajectories and impact the vane. Implications for engine operation in particulate-laden environments are discussed.

Committee: Jeffrey Bons PhD (Advisor); Micheal Dunn PhD (Committee Member) Subjects: Aerospace Engineering; Aerospace Materials; Engineering

Doctor of Philosophy, The Ohio State University, 2023, Mechanical Engineering

This thesis aimed to improve the numerical framework of modeling particle deposition in environments relevant to the cooling passage of the gas turbine engines. The effort comprised three main parts: (i) Investigating forces affecting particle motions. (ii) Applying smoothing techniques to improve convergence for mesh morphing. (iii) Elucidating experimental results through numerical simulations and sensitizing the OSU deposition model to temperature. Forces affecting particle trajectories were examined with a focus on turbophoretic force and Saffman lift force. Stochastic models were employed to generate the fluctuating flow velocity for modeling the turbophoretic force. Concentration and axial distributions of deposition velocity, as well as deposition velocity versus particle relaxation time, were predicted by The Discrete Random Walk model (DRW) and variations of the Continuous Random Walk model (CRW). These predictions were compared with existing experiments and direct numerical simulation (DNS) using two different geometries. Special attention was given to the inconsistent directional vector that results from solving the Langevin equation in non-rectangular geometries using location-specific coordinate systems. To achieve more physical solutions, the Langevin equation in cylindrical coordinates was derived for the pipe flows. The effect of the Saffman lift force with two random walk models was also discussed. To improve the stability of mesh morphing for modeling deposit growth, smoothing techniques were employed to reduce the irregular surface resulting from the discreteness of numerical setup and the effect of the Saffman lift force. Three schemes, the mass-based method proposed by Forsyth et al, the inverse distance weighting method and the radial basis function method, were employed to model the growth of the deposit. Because the scale of real cooling passages in gas turbine engines makes it difficult to use a fine mesh, simulations using both a fine and (open full item for complete abstract)

Committee: Jeffrey Bons Dr. (Advisor); Lian Duan Dr. (Committee Member); Randall Mathison Dr. (Committee Member) Subjects: Aerospace Engineering; Mechanical Engineering

Master of Science, The Ohio State University, 2016, Mechanical Engineering

The deposition of particulate in gas turbine cooling systems with a focus on single wall effusion holes was investigated. This study focused on the effect that flow turning angle into the cooling hole has on the blockage of these holes. The test hardware is based on a single walled combustor liner with angled effusion holes. By allowing the mass flow through the test system to decrease as deposition occurred the pressure drop across the test coupon was held at 3% of the discharge pressure. The mean flow turning angle was varied between favorable (10°) and adverse (130°) by mounting the plate in different orientations on a stalled plenum. The dust used was 0-10 µm Arizona Road Dust (ARD). These tests were run with a coupon temperature of 870 °C; this was achieved by use of an electric kiln. Flow reduction of the adverse test plates was around twice as much as the favorable condition; however both conditions had very similar capture efficiencies. 3D scans and sectioned test plates were used to investigate the different structures of the deposition that formed on the test plates and in the effusion holes. It is seen that turning angle does not influence the amount of captured mass but just the location of where that mass is captured and so its effect on the flow. A companion CFD study was also performed to explore the ability of computational models to predict the impact location and deposition depending on the impingement angle. This model was a simplified case and modeled a single effusion hole with the same geometry as the test plate. The inlet conditions were held constant and based on the experimental data. Particles were tracked with an Eulerian-Lagrangian method and it was seen that the predicted first impact locations closely matched the deposition seen in the experimental setup. Additionally a sticking model was used to predict deposition. It was seen that under the simulated conditions this model predicted deposition similar to the experimenta (open full item for complete abstract)

Committee: Jeffrey Bons Dr. (Advisor); Randall Mathison Dr. (Committee Member) Subjects: Aerospace Engineering; Fluid Dynamics; Mechanical Engineering

Doctor of Philosophy, The Ohio State University, 2013, Aero/Astro Engineering

Experimental and computational studies were conducted regarding particle deposition in the internal film cooling cavities of nozzle guide vanes. An experimental facility was fabricated to simulate particle deposition on an impingement liner and upstream surface of a nozzle guide vane wall. The facility supplied particle-laden flow at temperatures up to 1000°F (540°C) to a simplified impingement cooling test section. The heated flow passed through a perforated impingement plate and impacted on a heated flat wall. The particle-laden impingement jets resulted in the buildup of deposit cones associated with individual impingement jets. The deposit growth rate increased with increasing temperature and decreasing impinging velocities. For some low flow rates or high flow temperatures, the deposit cones heights spanned the entire gap between the impingement plate and wall, and grew through the impingement holes. For high flow rates, deposit structures were removed by shear forces from the flow. At low temperatures, deposit formed not only as individual cones, but as ridges located at the mid-planes between impinging jets. A computational model was developed to predict the deposit buildup seen in the experiments. The test section geometry and fluid flow from the experiment were replicated computationally and an Eulerian-Lagrangian particle tracking technique was employed. Several particle sticking models were employed and tested for adequacy. Sticking models that accurately predicted locations and rates in external deposition experiments failed to predict certain structures or rates seen in internal applications. A geometry adaptation technique was employed and the effect on deposition prediction was discussed. A new computational sticking model was developed that predicts deposition rates based on the local wall shear. The growth patterns were compared to experiments under different operating conditions. Of all the sticking models employed, the model based on wall shear, (open full item for complete abstract)

Committee: Jeffrey Bons (Advisor); Ali Ameri (Committee Member); Michael Dunn (Committee Member); Datta Gaitonde (Committee Member); Sandip Mazumder (Committee Member) Subjects: Aerospace Engineering